From steve.shuo.feng at gmail.com  Mon Jul  2 09:56:50 2012
From: steve.shuo.feng at gmail.com (Feng Shuo)
Date: Mon, 2 Jul 2012 15:56:50 +0800
Subject: [zeromq-dev] OpenSUSE Build Service for ZeroMQ
In-Reply-To: <CADL5_sgZESroQnsZRkKC=1LtQXv6-sVW8rxAg8df+N3Ax1eHbw@mail.gmail.com>
References: <CALmpHyYR7vEsBabixeTkD+NBATzaOxjuNLgyiDDq7aa=rRC6Yw@mail.gmail.com>
	<CALmpHya1n8KMkAFmFe2pUrWXsmNw61F9SHHXTjv1WxuB57n_Eg@mail.gmail.com>
	<CADL5_sgZESroQnsZRkKC=1LtQXv6-sVW8rxAg8df+N3Ax1eHbw@mail.gmail.com>
Message-ID: <CALmpHyYxUs2wdZSQEGrrx9hS8rKAOBGnh5WaWF3KF4ExgiMjsg@mail.gmail.com>

Hi Pieter,

I created the page at http://www.zeromq.org/distro:opensuse. Since I'm not
a native English speaker, I'm very happy if anyone could review the page
and fix any grammar issues. :-)

On Thu, Jun 28, 2012 at 4:04 PM, Pieter Hintjens <ph at imatix.com> wrote:

> On Wed, Jun 27, 2012 at 3:43 PM, Feng Shuo <steve.shuo.feng at gmail.com>
> wrote:
>
> > Just an update. I create two upstream builds in the OBS by it's "git
> source"
> > service. Whenever there is a commit, OBS will pull the updates and do a
> on
> > all platforms...
>
> Nice. On the http://www.zeromq.org/community page there is a section
> for builds and distros, and you might add an OpenSUSE page here
> (http://zeromq.distro:opensuse).
>
> -Pieter
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Feng Shuo
Tel: (86)10-59851155-2116
Fax: (86)10-59851155-2008
Tianjin Zhongke Blue Whale Information Technologies Co., Ltd
10th Floor, Tower A, The GATE building, No. 19 Zhong-guan-cun Avenue
Haidian District, Beijing, China
Postcode 100080
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120702/305e54b9/attachment.htm 

From ph at imatix.com  Mon Jul  2 12:02:32 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Mon, 2 Jul 2012 19:02:32 +0900
Subject: [zeromq-dev] OpenSUSE Build Service for ZeroMQ
In-Reply-To: <CALmpHyYxUs2wdZSQEGrrx9hS8rKAOBGnh5WaWF3KF4ExgiMjsg@mail.gmail.com>
References: <CALmpHyYR7vEsBabixeTkD+NBATzaOxjuNLgyiDDq7aa=rRC6Yw@mail.gmail.com>
	<CALmpHya1n8KMkAFmFe2pUrWXsmNw61F9SHHXTjv1WxuB57n_Eg@mail.gmail.com>
	<CADL5_sgZESroQnsZRkKC=1LtQXv6-sVW8rxAg8df+N3Ax1eHbw@mail.gmail.com>
	<CALmpHyYxUs2wdZSQEGrrx9hS8rKAOBGnh5WaWF3KF4ExgiMjsg@mail.gmail.com>
Message-ID: <CADL5_si8cLKbMisA-XJgb4LUza8+e8Cd_Net+tWDX5+CNObVJg@mail.gmail.com>

Steve,

Looks great, and I didn't see a single spelling or grammar issue... :)

-Pieter

On Mon, Jul 2, 2012 at 4:56 PM, Feng Shuo <steve.shuo.feng at gmail.com> wrote:

> I created the page at http://www.zeromq.org/distro:opensuse. Since I'm not a
> native English speaker, I'm very happy if anyone could review the page and
> fix any grammar issues. :-)

From dacmorton at gmail.com  Mon Jul  2 17:09:15 2012
From: dacmorton at gmail.com (Daniel Morton)
Date: Mon, 2 Jul 2012 15:09:15 +0000 (UTC)
Subject: [zeromq-dev] Make Install Error
Message-ID: <loom.20120702T170557-129@post.gmane.org>

I run ./configure and make seemingly without incident; 
make check passes all 9 tests but when I run make 
install I get the following error message:


Making install in src
test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
/bin/sh ../libtool --mode=install /usr/bin/install -c  libzmq.la '/usr/local/lib'
libtool: install:
/usr/bin/install -c .libs/libzmq.1.dylib /usr/local/lib/libzmq.1.dylib
install: /usr/local/lib/libzmq.1.dylib: Permission denied
make[2]: *** [install-libLTLIBRARIES] Error 71
make[1]: *** [install-am] Error 2
make: *** [install-recursive] Error 1

What isn't working?


From pelletier.michel at gmail.com  Mon Jul  2 17:45:59 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 2 Jul 2012 08:45:59 -0700
Subject: [zeromq-dev] Make Install Error
In-Reply-To: <loom.20120702T170557-129@post.gmane.org>
References: <loom.20120702T170557-129@post.gmane.org>
Message-ID: <CACxu=vLeojQs-rYUMP44eZHWtGYvWr81JO_Hx_e46uJWyXPjYw@mail.gmail.com>

By default make install installs to /usr/local which requires root permissions.

You can either run make install as root or pass
"--prefix=/different/path" to configure to install it to a different
path.

-Michel

On Mon, Jul 2, 2012 at 8:09 AM, Daniel Morton <dacmorton at gmail.com> wrote:
> I run ./configure and make seemingly without incident;
> make check passes all 9 tests but when I run make
> install I get the following error message:
>
>
> Making install in src
> test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
> /bin/sh ../libtool --mode=install /usr/bin/install -c  libzmq.la '/usr/local/lib'
> libtool: install:
> /usr/bin/install -c .libs/libzmq.1.dylib /usr/local/lib/libzmq.1.dylib
> install: /usr/local/lib/libzmq.1.dylib: Permission denied
> make[2]: *** [install-libLTLIBRARIES] Error 71
> make[1]: *** [install-am] Error 2
> make: *** [install-recursive] Error 1
>
> What isn't working?
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From minling.ann at gmail.com  Mon Jul  2 12:19:20 2012
From: minling.ann at gmail.com (=?Big5?B?qvSx07lh?=)
Date: Mon, 2 Jul 2012 18:19:20 +0800
Subject: [zeromq-dev] zeromq epgm transport cannot work
Message-ID: <CAOmGT9R=Y=6chGc1h644Lua179gFi4hKhdQetvXqQJ1QEqas3w@mail.gmail.com>

hi~

There are some problems when i use the zeromq epgm transport.

My system is ubuntu 12.04.

I have built the source with --with-pgm option.

Problem 1:

zmq_bind(socket, "epgm://eth0;224.0.0.1:5555");
zmq_connect(socket, "epgm://eth0;224.0.0.1:5555");

even i place the eth0 with it's ip address but both sides have error
message:

Assertion failed: res != NULL (pgm_socket.cpp:137)
Abort (core dumped)

Problem 2:

zmq_bind(socket, "epgm://*;224.0.0.1:5555");
zmq_connect(socket, "epgm://*;224.0.0.1:5555")

There are no error message but the receiver side doesn't receive any
messages.

Please help me. I don't know what's wrong.

Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120702/003158a8/attachment.htm 

From steven.mccoy at miru.hk  Mon Jul  2 19:13:36 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Mon, 2 Jul 2012 13:13:36 -0400
Subject: [zeromq-dev] zeromq epgm transport cannot work
In-Reply-To: <CAOmGT9R=Y=6chGc1h644Lua179gFi4hKhdQetvXqQJ1QEqas3w@mail.gmail.com>
References: <CAOmGT9R=Y=6chGc1h644Lua179gFi4hKhdQetvXqQJ1QEqas3w@mail.gmail.com>
Message-ID: <CAGFWLihGuOFzQuNFrM5m-2=2xadN=+5rJT7XYwMieR6-pCVP8w@mail.gmail.com>

On Monday, 2 July 2012, ??? wrote:

> hi~
>
> There are some problems when i use the zeromq epgm transport.
>
> My system is ubuntu 12.04.
>
> I have built the source with --with-pgm option.
>
> Problem 1:
>
> zmq_bind(socket, "epgm://eth0;224.0.0.1:5555");
> zmq_connect(socket, "epgm://eth0;224.0.0.1:5555");
>
> even i place the eth0 with it's ip address but both sides have error
> message:
>
> Assertion failed: res != NULL (pgm_socket.cpp:137)
> Abort (core dumped)
>
> Problem 2:
>
> zmq_bind(socket, "epgm://*;224.0.0.1:5555");
> zmq_connect(socket, "epgm://*;224.0.0.1:5555")
>
> There are no error message but the receiver side doesn't receive any
> messages.
>
> Please help me. I don't know what's wrong.
>
> Thanks.
>

If this is the 3.x release candidate please try git head instead.  This is
a known and resolved defect.

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120702/2376fdc5/attachment.htm 

From lists at chuckremes.com  Mon Jul  2 20:48:25 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 2 Jul 2012 13:48:25 -0500
Subject: [zeromq-dev] daily build cluster question
Message-ID: <96219A52-FDEA-4377-9ED4-6C17B00C0E6B@chuckremes.com>

I see on the #zeromq.build channel that the ffi-rzmq gem has failures on Debian linux. I'm the maintainer of that gem and would like to fix those failures. On my OSX box *and* my Archlinux box, the specs all pass. What version of Debian Linux (3, 4, 5 or 6) should I install to match the build cluster so I can reproduce this problem?

cr


From tolitius at gmail.com  Mon Jul  2 20:48:09 2012
From: tolitius at gmail.com (Anatoly)
Date: Mon, 2 Jul 2012 14:48:09 -0400
Subject: [zeromq-dev] Releasing RAM in Chunks
Message-ID: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>

?MQ Crowd,

    Reading messages from a socket, placing them on the zmq (PUSH). On the other side reading messages of off the queue (PULL) and persisting them in to DB.  

    If we get millions of messages, ?MQ takes X GB of RAM (since the pushing in this case is at much higher speed than pulling that waits for a DB for each pull), and does not release these Gigs until ALL of the messages are consumed (e.g. pulled).  

    Is there a way to configure it to release memory in chunks as the queue is being emptied?

Thank you,
/Anatoly

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120702/410882f6/attachment.htm 

From lists at chuckremes.com  Mon Jul  2 20:53:52 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 2 Jul 2012 13:53:52 -0500
Subject: [zeromq-dev] Releasing RAM in Chunks
In-Reply-To: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>
References: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>
Message-ID: <5497E2E4-701F-485D-9E1D-EFB9AA242BD3@chuckremes.com>

On Jul 2, 2012, at 1:48 PM, Anatoly wrote:

> ?MQ Crowd,
> 
>     Reading messages from a socket, placing them on the zmq (PUSH). On the other side reading messages of off the queue (PULL) and persisting them in to DB. 
> 
>     If we get millions of messages, ?MQ takes X GB of RAM (since the pushing in this case is at much higher speed than pulling that waits for a DB for each pull), and does not release these Gigs until ALL of the messages are consumed (e.g. pulled). 
> 
>     Is there a way to configure it to release memory in chunks as the queue is being emptied?

What you have posted should not be true. The memory for a message will be released when you call zmq_msg_close(). If you are making that call and the memory isn't freed, then there is a bug. See the man page at http://api.zeromq.org/2-2:zmq-msg-close

If you see something different, then please provide an example (written in C) that demonstrates the problem and open a bug. Thank you.

cr



From justin at affinix.com  Mon Jul  2 21:57:33 2012
From: justin at affinix.com (Justin Karneges)
Date: Mon, 2 Jul 2012 12:57:33 -0700
Subject: [zeromq-dev] HWM behaviour & blocking
In-Reply-To: <CAA0gF6qgduRbrFrfBJuZa75ViVKO+tDXi6C5u_1viRdLQYMWCA@mail.gmail.com>
References: <4FAA0461.1050205@steffen-mueller.net>
	<201206281106.38406.justin@affinix.com>
	<CAA0gF6qgduRbrFrfBJuZa75ViVKO+tDXi6C5u_1viRdLQYMWCA@mail.gmail.com>
Message-ID: <201207021257.33721.justin@affinix.com>

On Friday, June 29, 2012 06:13:53 AM Paul Colomiets wrote:
> On Thu, Jun 28, 2012 at 9:06 PM, Justin Karneges <justin at affinix.com> wrote:
> > It's really just for functional completeness of my event-driven wrapper.
> > The only time I can see this coming up in practice is an application
> > that pushes a message just before exiting.
> > 
> > For now, I set ZMQ_LINGER to 0 when a socket object is destroyed, making
> > the above application impossible to create. What I'm thinking of doing
> > now is offering an alternate, blocking-based shutdown method. This would
> > violate the spirit of my wrapper, but may work well enough for apps that
> > finish with a single socket doing a write-and-exit.
> 
> I think you should just set linger and use it. zmq_close() doesn't
> block. The zmq_term() blocks.

Wow, silly me working around a non-problem. I was assuming zmq_close() 
blocked. Thanks for clarifying.

Justin

From ph at imatix.com  Tue Jul  3 02:02:38 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Tue, 3 Jul 2012 09:02:38 +0900
Subject: [zeromq-dev] Releasing RAM in Chunks
In-Reply-To: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>
References: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>
Message-ID: <CADL5_shT5aF2rFpmR7cb-bMX0=wsjY6ZmByO6icuK=FeZQCXSQ@mail.gmail.com>

Anatoly,

The standard causes for memory exhaustion are, in order:

* You're forgetting to close messages, or doing it at the wrong time.
* Your consumer is unable to process messages at full speed, so the
queue builds up.

In the second case you need to decide how you want to handle the excess.

-Pieter

On Tue, Jul 3, 2012 at 3:48 AM, Anatoly <tolitius at gmail.com> wrote:
> ?MQ Crowd,
>
>     Reading messages from a socket, placing them on the zmq (PUSH). On the
> other side reading messages of off the queue (PULL) and persisting them in
> to DB.
>
>     If we get millions of messages, ?MQ takes X GB of RAM (since the pushing
> in this case is at much higher speed than pulling that waits for a DB for
> each pull), and does not release these Gigs until ALL of the messages are
> consumed (e.g. pulled).
>
>     Is there a way to configure it to release memory in chunks as the queue
> is being emptied?
>
> Thank you,
> /Anatoly
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From steve.shuo.feng at gmail.com  Tue Jul  3 08:40:33 2012
From: steve.shuo.feng at gmail.com (Feng Shuo)
Date: Tue, 3 Jul 2012 14:40:33 +0800
Subject: [zeromq-dev] OpenSUSE Build Service for ZeroMQ
In-Reply-To: <CADL5_si8cLKbMisA-XJgb4LUza8+e8Cd_Net+tWDX5+CNObVJg@mail.gmail.com>
References: <CALmpHyYR7vEsBabixeTkD+NBATzaOxjuNLgyiDDq7aa=rRC6Yw@mail.gmail.com>
	<CALmpHya1n8KMkAFmFe2pUrWXsmNw61F9SHHXTjv1WxuB57n_Eg@mail.gmail.com>
	<CADL5_sgZESroQnsZRkKC=1LtQXv6-sVW8rxAg8df+N3Ax1eHbw@mail.gmail.com>
	<CALmpHyYxUs2wdZSQEGrrx9hS8rKAOBGnh5WaWF3KF4ExgiMjsg@mail.gmail.com>
	<CADL5_si8cLKbMisA-XJgb4LUza8+e8Cd_Net+tWDX5+CNObVJg@mail.gmail.com>
Message-ID: <CALmpHyYb-bL-3DR8RKSkkyA506bOn84DXsQ6tAcYt7865QNbtQ@mail.gmail.com>

Thank you. I also added pages for RHEL/CentOS and SLES.
http://www.zeromq.org/distro:centos
and
http://www.zeromq.org/distro:sles

On Mon, Jul 2, 2012 at 6:02 PM, Pieter Hintjens <ph at imatix.com> wrote:

> Steve,
>
> Looks great, and I didn't see a single spelling or grammar issue... :)
>
> -Pieter
>
> On Mon, Jul 2, 2012 at 4:56 PM, Feng Shuo <steve.shuo.feng at gmail.com>
> wrote:
>
> > I created the page at http://www.zeromq.org/distro:opensuse. Since I'm
> not a
> > native English speaker, I'm very happy if anyone could review the page
> and
> > fix any grammar issues. :-)
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Feng Shuo
Tel: (86)10-59851155-2116
Fax: (86)10-59851155-2008
Tianjin Zhongke Blue Whale Information Technologies Co., Ltd
10th Floor, Tower A, The GATE building, No. 19 Zhong-guan-cun Avenue
Haidian District, Beijing, China
Postcode 100080
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120703/089eec34/attachment.htm 

From paul at colomiets.name  Tue Jul  3 10:25:26 2012
From: paul at colomiets.name (Paul Colomiets)
Date: Tue, 3 Jul 2012 11:25:26 +0300
Subject: [zeromq-dev] Releasing RAM in Chunks
In-Reply-To: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>
References: <DCB5C83623EB4FDC887A3DF732B3AC27@gmail.com>
Message-ID: <CAA0gF6qh1pTQhNRZGKFtf+S0tAwMMQvAq5uLdstVP1wT-sJ0bw@mail.gmail.com>

Hi Anatoly,

On Mon, Jul 2, 2012 at 9:48 PM, Anatoly <tolitius at gmail.com> wrote:
> ?MQ Crowd,
>
>     Reading messages from a socket, placing them on the zmq (PUSH). On the
> other side reading messages of off the queue (PULL) and persisting them in
> to DB.
>
>     If we get millions of messages, ?MQ takes X GB of RAM (since the pushing
> in this case is at much higher speed than pulling that waits for a DB for
> each pull), and does not release these Gigs until ALL of the messages are
> consumed (e.g. pulled).
>
>     Is there a way to configure it to release memory in chunks as the queue
> is being emptied?
>

Unless you do something wrong, as others described. The effect you see
may be because of memory fragmentation, so as last resort you can try
to link your app against jemalloc, as there is evidence that jemalloc
handles fragmentation better.

In any case please report what fixed the problem.

-- 
Paul

From rblists at gmail.com  Wed Jul  4 12:05:53 2012
From: rblists at gmail.com (Raphael Bauduin)
Date: Wed, 4 Jul 2012 12:05:53 +0200
Subject: [zeromq-dev] Interrupted System Call: advice to handle it
Message-ID: <CAONrwUHbDUGwiOz-_Qc81m+Vn+uninnUjioSar8_NDO4JZTH3w@mail.gmail.com>

Hi,

I'm using the ruby zmq bindings in a web application. I regularly get
error message "ZMQ::Error: Interrupted system call" related to a send.
This is in a Ruby on Rails application served with passenger, which
spawns worker processes. I think I have identified a process that
generated this error, and an strace on it shows no activity at all.
This process however keeps open a connection to the mysql server. An
accumulation of such errors will eventually become problematic server
side, in addition to clients getting an error page and messages being
lost.


I'm looking for advice in avoiding this error and possibly for further
debugging hints. Related to that I have several questions:
- Should I simply catch this exception, and retry the send if needed?
As this is done in the process sending the page content back to the
client, won't it possibly make some requests too slow? (This could
still be better than an error as we have currently)
- If my understanding is correct, the problem occurs with blocking
syscalls, and requests having the error don't return any content to
the client. But what happens if I make the send non blocking?
(http://zeromq.github.com/rbzmq/classes/ZMQ/Socket.html#M000010)
- Finally, what might interrupt the syscall? Any interesting read about this?

Thanks in advance

Raph

From krikun.daniel at gmail.com  Wed Jul  4 12:22:08 2012
From: krikun.daniel at gmail.com (Daniel Krikun)
Date: Wed, 4 Jul 2012 13:22:08 +0300
Subject: [zeromq-dev] match request and reply
Message-ID: <CAMohcw1vCs5Fgbt=LEuo+T9S9QORoNX7YbZZDWDNge0jf4XbKg@mail.gmail.com>

Hello,

I'm building a service-oriented system. A client sends a request to create
some entity on the server. The server creates the entity and then replies
with the entity's handle, so that the client can send more requests
regarding the entity (i.e. modify/delete it).
I need to match the request to create an entity with the reply containing
its handle (otherwise a client could potentially confuse between different
entities' handles).

Is there any recommended way do that in zeromq? Any 'best practices'?

If I understand that correctly, using REQ-REP socket combination ensures
that the reply will match the request, however:
1. The server is multi-threaded and does not ensures
first-come-first-served policy
2. Some request are left without reply intentionally, which is not possible
with REQ-REP, so I actually use DEALER-DEALER combination.

Regards,

-- 
Daniel Krikun
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120704/1c2df728/attachment.htm 

From lists at chuckremes.com  Wed Jul  4 17:16:11 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Wed, 4 Jul 2012 10:16:11 -0500
Subject: [zeromq-dev] match request and reply
In-Reply-To: <CAMohcw1vCs5Fgbt=LEuo+T9S9QORoNX7YbZZDWDNge0jf4XbKg@mail.gmail.com>
References: <CAMohcw1vCs5Fgbt=LEuo+T9S9QORoNX7YbZZDWDNge0jf4XbKg@mail.gmail.com>
Message-ID: <84DCE31A-B237-4DC9-879F-AEA0A819D3BF@chuckremes.com>


On Jul 4, 2012, at 5:22 AM, Daniel Krikun wrote:

> Hello,
> 
> I'm building a service-oriented system. A client sends a request to create some entity on the server. The server creates the entity and then replies with the entity's handle, so that the client can send more requests regarding the entity (i.e. modify/delete it).
> I need to match the request to create an entity with the reply containing its handle (otherwise a client could potentially confuse between different entities' handles).
> 
> Is there any recommended way do that in zeromq? Any 'best practices'?

I usually use the Majordomo protocol / pattern as the basis for most of the work I do.

> If I understand that correctly, using REQ-REP socket combination ensures that the reply will match the request, however: 
> 1. The server is multi-threaded and does not ensures first-come-first-served policy

You solve this by using Dealer/Router.

> 2. Some request are left without reply intentionally, which is not possible with REQ-REP, so I actually use DEALER-DEALER combination.

You should use Dealer/ROUTER instead of Dealer/Dealer. 

cr



From lists at chuckremes.com  Wed Jul  4 17:20:01 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Wed, 4 Jul 2012 10:20:01 -0500
Subject: [zeromq-dev] Interrupted System Call: advice to handle it
In-Reply-To: <CAONrwUHbDUGwiOz-_Qc81m+Vn+uninnUjioSar8_NDO4JZTH3w@mail.gmail.com>
References: <CAONrwUHbDUGwiOz-_Qc81m+Vn+uninnUjioSar8_NDO4JZTH3w@mail.gmail.com>
Message-ID: <01B73EA7-9C85-467A-AF3E-533BC52E5945@chuckremes.com>


On Jul 4, 2012, at 5:05 AM, Raphael Bauduin wrote:

> Hi,
> 
> I'm using the ruby zmq bindings in a web application. I regularly get
> error message "ZMQ::Error: Interrupted system call" related to a send.
> This is in a Ruby on Rails application served with passenger, which
> spawns worker processes. I think I have identified a process that
> generated this error, and an strace on it shows no activity at all.
> This process however keeps open a connection to the mysql server. An
> accumulation of such errors will eventually become problematic server
> side, in addition to clients getting an error page and messages being
> lost.

I'm assuming this happens under MRI. Is it 1.8.x or 1.9.x?

Do you see the same behavior when running your app with JRuby or Rubinius?

> I'm looking for advice in avoiding this error and possibly for further
> debugging hints. Related to that I have several questions:
> - Should I simply catch this exception, and retry the send if needed?
> As this is done in the process sending the page content back to the
> client, won't it possibly make some requests too slow? (This could
> still be better than an error as we have currently)

Using exception handling for flow control in Ruby can be slow. But unless you are building the next amazon.com then it probably won't hurt you too much. You could give this a try though it's always better to figure out the actual underlying cause and fix it. Using exceptions here is just a band-aid.

> - If my understanding is correct, the problem occurs with blocking
> syscalls, and requests having the error don't return any content to
> the client. But what happens if I make the send non blocking?
> (http://zeromq.github.com/rbzmq/classes/ZMQ/Socket.html#M000010)

Try it and see.

> - Finally, what might interrupt the syscall? Any interesting read about this?

Something in your app is generating a signal. The technique I use to figure out these kinds of errors is to run my app under other Ruby runtimes. Most of the time they will fail differently and/or give me an exact backtrace pointing to the source of the problem.

Lastly, you may want to look at the ffi-rzmq gem (disclaimer: I'm its maintainer). It has a different API from the zmq gem but it appears to enjoy wider usage by the community so it may be a bit more stable.

cr


From mark.farnan at petrolink.com  Wed Jul  4 20:30:29 2012
From: mark.farnan at petrolink.com (Mark Farnan)
Date: Wed, 4 Jul 2012 13:30:29 -0500
Subject: [zeromq-dev] Websockets as a Transport ?
Message-ID: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>

Howdy all,

 

I'm evaluating zeroMQ, and one main criteria we have is to support
Websockets for client to server communications.

 

Has there been any work on making a native Websocket transport for ZeroMQ in
the core C++ libraries ?

 

 

Regards

 

Mark Farnan

Director, Product Development

Petrolink Services Inc

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120704/d5b42ddd/attachment.htm 

From xekoukou at gmail.com  Wed Jul  4 22:37:23 2012
From: xekoukou at gmail.com (Apostolis Xekoukoulotakis)
Date: Wed, 4 Jul 2012 23:37:23 +0300
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
Message-ID: <CAOX4E5G09FEASKvWYhvFFLEwzcEsN_DiaBhC34Bj-GmN8U5Q+Q@mail.gmail.com>

This isnt the official response.

It is very easy though to create a bridge between the two if you use nodejs
and the socket.io module.

There was also a project to put the zeromq semantics in the browser.
https://github.com/progrium/nullmq


2012/7/4 Mark Farnan <mark.farnan at petrolink.com>

> Howdy all,****
>
> ** **
>
> I?m evaluating zeroMQ, and one main criteria we have is to support
> Websockets for client to server communications.****
>
> ** **
>
> Has there been any work on making a native Websocket transport for ZeroMQ
> in the core C++ libraries ?****
>
> ** **
>
> ** **
>
> Regards****
>
> ** **
>
> Mark Farnan****
>
> Director, Product Development****
>
> Petrolink Services Inc****
>
> ** **
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


-- 


Sincerely yours,

     Apostolis Xekoukoulotakis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120704/77b7a2ff/attachment.htm 

From mark.farnan at petrolink.com  Wed Jul  4 23:06:51 2012
From: mark.farnan at petrolink.com (Mark Farnan)
Date: Wed, 4 Jul 2012 16:06:51 -0500
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <CAOX4E5G09FEASKvWYhvFFLEwzcEsN_DiaBhC34Bj-GmN8U5Q+Q@mail.gmail.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
	<CAOX4E5G09FEASKvWYhvFFLEwzcEsN_DiaBhC34Bj-GmN8U5Q+Q@mail.gmail.com>
Message-ID: <028901cd5a28$ef548da0$cdfda8e0$@petrolink.com>

Thanks for the quick reply.

 

I am actually not so much interested in the browser, as I am in using
Websockets for the underlying transport between applications / clients &
Server, which have to communicate across the Internet,  (over firewalls).

 

Websockets works extremely well for this purpose, to provide a bidirectional
socket based connection, over ports 80/443.  Many corporate firewalls have
TCP ports blocked, and opening them is not an option for our application.

 

ZeroMQ looks perfect for the task in all things, except no Websocket
transport (to replace the TCP transport), so I was hoping someone has looked
at it already.

Writing it ourselves may be an option,  vs using another of the Messaging
systems.

 

Regards

 

Mark Farnan

 

From: zeromq-dev-bounces at lists.zeromq.org
[mailto:zeromq-dev-bounces at lists.zeromq.org] On Behalf Of Apostolis
Xekoukoulotakis
Sent: Wednesday, July 04, 2012 3:37 PM
To: ZeroMQ development list
Subject: Re: [zeromq-dev] Websockets as a Transport ?

 

This isnt the official response. 

It is very easy though to create a bridge between the two if you use nodejs
and the socket.io module.

There was also a project to put the zeromq semantics in the browser.
https://github.com/progrium/nullmq



2012/7/4 Mark Farnan <mark.farnan at petrolink.com>

Howdy all,

 

I'm evaluating zeroMQ, and one main criteria we have is to support
Websockets for client to server communications.

 

Has there been any work on making a native Websocket transport for ZeroMQ in
the core C++ libraries ?

 

 

Regards

 

Mark Farnan

Director, Product Development

Petrolink Services Inc

 


_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev




-- 



 
Sincerely yours, 
     Apostolis Xekoukoulotakis

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120704/0185cf93/attachment.htm 

From ian.barber at gmail.com  Wed Jul  4 23:44:58 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Wed, 4 Jul 2012 22:44:58 +0100
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
Message-ID: <CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>

On Wed, Jul 4, 2012 at 7:30 PM, Mark Farnan <mark.farnan at petrolink.com>wrote:

> Howdy all,****
>
> ** **
>
> I?m evaluating zeroMQ, and one main criteria we have is to support
> Websockets for client to server communications.****
>
> ** **
>
> Has there been any work on making a native Websocket transport for ZeroMQ
> in the core C++ libraries ?
>

No work that made it anywhere near head, but as Apostolis said there are
plenty of bridging methods. My personal favourite is Paul's ZeroGW:
https://github.com/tailhook/zerogw

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120704/76e75e0d/attachment.htm 

From cangussu at gmail.com  Thu Jul  5 00:24:21 2012
From: cangussu at gmail.com (Thiago Cangussu)
Date: Wed, 4 Jul 2012 19:24:21 -0300
Subject: [zeromq-dev] How to safely use a socket in multiple threads in zmq
	2.1.11
Message-ID: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>

Hi,

I have a pool of threads that needs to send messages to a zmq socket. I was
creating the socket, connecting, sending the message and closing it for
each message, but it is for sure not an optimal usage of resources, since
if I create/close sockets too fast, zmq is not able do release them as fast
and I get the error: "too many open files (signaler.cpp 330)". Since I have
no control over those threads, I'm about to try the following solution:

1 - create a pool of connected sockets (REQ, inproc)
2 - when a client thread wants to send a message, it grabs a socket from
the pool
3 - the thread performs a send/recv
4 - the thread releases the socket (which will be again available in the
pool)

The question is: what should I do to make this multi-threaded usage of
sockets safe? Is it enough to guarantee that multiple threads do not
perform send/receive simultaneously?

Thanks,
Thiago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120704/dba017b0/attachment.htm 

From shemminger at vyatta.com  Thu Jul  5 00:57:15 2012
From: shemminger at vyatta.com (Stephen Hemminger)
Date: Wed, 4 Jul 2012 15:57:15 -0700
Subject: [zeromq-dev] How to safely use a socket in multiple threads in
 zmq 2.1.11
In-Reply-To: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
References: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
Message-ID: <20120704155715.63ffadfe@nehalam.linuxnetplumber.net>

On Wed, 4 Jul 2012 19:24:21 -0300
Thiago Cangussu <cangussu at gmail.com> wrote:

> Hi,
> 
> I have a pool of threads that needs to send messages to a zmq socket. I was
> creating the socket, connecting, sending the message and closing it for
> each message, but it is for sure not an optimal usage of resources, since
> if I create/close sockets too fast, zmq is not able do release them as fast
> and I get the error: "too many open files (signaler.cpp 330)". Since I have
> no control over those threads, I'm about to try the following solution:
> 
> 1 - create a pool of connected sockets (REQ, inproc)
> 2 - when a client thread wants to send a message, it grabs a socket from
> the pool
> 3 - the thread performs a send/recv
> 4 - the thread releases the socket (which will be again available in the
> pool)
> 
> The question is: what should I do to make this multi-threaded usage of
> sockets safe? Is it enough to guarantee that multiple threads do not
> perform send/receive simultaneously?
> 
> Thanks,
> Thiago

You can have a per-thread pool of sockets, but you can't create socket
in one thread and then use it in another. Can you just have one socket
per thread in thread local storage?

From whjackson at gmail.com  Thu Jul  5 02:57:36 2012
From: whjackson at gmail.com (Whitney Jackson)
Date: Wed, 4 Jul 2012 19:57:36 -0500
Subject: [zeromq-dev] blocking zmq_close?
Message-ID: <CAJq74gGO0k6syZzTcK3oFQARmDWLVurT96-EvRMV5zb4Z7tc4Q@mail.gmail.com>

Hi,

I'm using the ZMQ_FD getsockopt feature to integrate ZeroMQ into an
event loop that I don't control and don't have the source code for.
This works quite well except when I want to close a ZeroMQ socket.
Since zmq_close is non-blocking, control gets returned to the event
loop before the socket is fully shutdown causing this error:

Bad file descriptor (epoll.cpp:69)

gdb on the core file shows

#0  0xf77bf430 in __kernel_vsyscall ()
#1  0x45d7395f in raise () from /lib/libc.so.6
#2  0x45d752b3 in abort () from /lib/libc.so.6
#3  0xf36f719c in zmq::zmq_abort (errmsg_=0x45eaf393 "Bad file
descriptor") at err.cpp:76
#4  0xf36f5f0a in zmq::epoll_t::add_fd (this=0x9029a10, fd_=14,
events_=0x902a2b0) at epoll.cpp:69
#5  0xf370c3d7 in zmq::socket_base_t::start_reaping (this=0x902a210,
poller_=0x9029a10) at socket_base.cpp:717
#6  0xf3705a87 in zmq::reaper_t::process_reap (this=0x90287e0,
socket_=0x902a210) at reaper.cpp:101
#7  0xf36fbfd4 in zmq::object_t::process_command (this=0x90287e0,
cmd_=...) at object.cpp:120
#8  0xf370591f in zmq::reaper_t::in_event (this=0x90287e0) at reaper.cpp:72
#9  0xf36f64f7 in zmq::epoll_t::loop (this=0x9029a10) at epoll.cpp:161
#10 0xf36f65c5 in zmq::epoll_t::worker_routine (arg_=0x9029a10) at epoll.cpp:174
#11 0xf3710792 in thread_routine (arg_=0x9029a54) at thread.cpp:75
#12 0xf7749adf in start_thread () from /lib/libpthread.so.0
#13 0x45e3655e in clone () from /lib/libc.so.6

I can prevent the error by doing a zmq_term on the context or just
sleeping before returning control to the event loop but obviously
these solutions aren't optimal.

Is there some way to block until zmq_close is really done?

Whitney

From justin at affinix.com  Thu Jul  5 03:52:11 2012
From: justin at affinix.com (Justin Karneges)
Date: Wed, 4 Jul 2012 18:52:11 -0700
Subject: [zeromq-dev] blocking zmq_close?
In-Reply-To: <CAJq74gGO0k6syZzTcK3oFQARmDWLVurT96-EvRMV5zb4Z7tc4Q@mail.gmail.com>
References: <CAJq74gGO0k6syZzTcK3oFQARmDWLVurT96-EvRMV5zb4Z7tc4Q@mail.gmail.com>
Message-ID: <201207041852.12047.justin@affinix.com>

On Wednesday, July 04, 2012 05:57:36 PM Whitney Jackson wrote:
> Hi,
> 
> I'm using the ZMQ_FD getsockopt feature to integrate ZeroMQ into an
> event loop that I don't control and don't have the source code for.
> This works quite well except when I want to close a ZeroMQ socket.
> Since zmq_close is non-blocking, control gets returned to the event
> loop before the socket is fully shutdown causing this error:
> 
> Bad file descriptor (epoll.cpp:69)
> 
> gdb on the core file shows
> 
> #0  0xf77bf430 in __kernel_vsyscall ()
> #1  0x45d7395f in raise () from /lib/libc.so.6
> #2  0x45d752b3 in abort () from /lib/libc.so.6
> #3  0xf36f719c in zmq::zmq_abort (errmsg_=0x45eaf393 "Bad file
> descriptor") at err.cpp:76
> #4  0xf36f5f0a in zmq::epoll_t::add_fd (this=0x9029a10, fd_=14,
> events_=0x902a2b0) at epoll.cpp:69
> #5  0xf370c3d7 in zmq::socket_base_t::start_reaping (this=0x902a210,
> poller_=0x9029a10) at socket_base.cpp:717
> #6  0xf3705a87 in zmq::reaper_t::process_reap (this=0x90287e0,
> socket_=0x902a210) at reaper.cpp:101
> #7  0xf36fbfd4 in zmq::object_t::process_command (this=0x90287e0,
> cmd_=...) at object.cpp:120
> #8  0xf370591f in zmq::reaper_t::in_event (this=0x90287e0) at reaper.cpp:72
> #9  0xf36f64f7 in zmq::epoll_t::loop (this=0x9029a10) at epoll.cpp:161
> #10 0xf36f65c5 in zmq::epoll_t::worker_routine (arg_=0x9029a10) at
> epoll.cpp:174 #11 0xf3710792 in thread_routine (arg_=0x9029a54) at
> thread.cpp:75 #12 0xf7749adf in start_thread () from /lib/libpthread.so.0
> #13 0x45e3655e in clone () from /lib/libc.so.6
> 
> I can prevent the error by doing a zmq_term on the context or just
> sleeping before returning control to the event loop but obviously
> these solutions aren't optimal.
> 
> Is there some way to block until zmq_close is really done?

How about removing the socket's fd from the eventloop's monitoring at the time 
of close? It would be surprising if you can add an fd but not remove it...

Justin

From whjackson at gmail.com  Thu Jul  5 04:15:09 2012
From: whjackson at gmail.com (Whitney Jackson)
Date: Wed, 4 Jul 2012 21:15:09 -0500
Subject: [zeromq-dev] blocking zmq_close?
In-Reply-To: <201207041852.12047.justin@affinix.com>
References: <CAJq74gGO0k6syZzTcK3oFQARmDWLVurT96-EvRMV5zb4Z7tc4Q@mail.gmail.com>
	<201207041852.12047.justin@affinix.com>
Message-ID: <CAJq74gFrW=SFsJz9fNpZ82kMvD1FhR0gXMp1hgKmk9FL4xKgRQ@mail.gmail.com>

> How about removing the socket's fd from the eventloop's monitoring at the time
> of close? It would be surprising if you can add an fd but not remove it...

Yep.  The propriety event loop does provide a mechanism for removing
fds as well as adding them. I use that and can verify using strace
that the event loop no longer includes my fd in it's select calls
afterwards.  Sadly, it still blows up as soon as I close the socket if
I don't zmq_term or sleep.  I'm guessing somewhere in the code I can't
see the fd is being touched in a way that doesn't show up in strace.

I could go the route of filing a bug with the commercial / closed
source event loop people but am guessing I won't have much luck there.
 If it's possible to work around this by just not giving control back
to the event loop until the socket is fully closed that would be
ideal.

On Wed, Jul 4, 2012 at 8:52 PM, Justin Karneges <justin at affinix.com> wrote:
> On Wednesday, July 04, 2012 05:57:36 PM Whitney Jackson wrote:
>> Hi,
>>
>> I'm using the ZMQ_FD getsockopt feature to integrate ZeroMQ into an
>> event loop that I don't control and don't have the source code for.
>> This works quite well except when I want to close a ZeroMQ socket.
>> Since zmq_close is non-blocking, control gets returned to the event
>> loop before the socket is fully shutdown causing this error:
>>
>> Bad file descriptor (epoll.cpp:69)
>>
>> gdb on the core file shows
>>
>> #0  0xf77bf430 in __kernel_vsyscall ()
>> #1  0x45d7395f in raise () from /lib/libc.so.6
>> #2  0x45d752b3 in abort () from /lib/libc.so.6
>> #3  0xf36f719c in zmq::zmq_abort (errmsg_=0x45eaf393 "Bad file
>> descriptor") at err.cpp:76
>> #4  0xf36f5f0a in zmq::epoll_t::add_fd (this=0x9029a10, fd_=14,
>> events_=0x902a2b0) at epoll.cpp:69
>> #5  0xf370c3d7 in zmq::socket_base_t::start_reaping (this=0x902a210,
>> poller_=0x9029a10) at socket_base.cpp:717
>> #6  0xf3705a87 in zmq::reaper_t::process_reap (this=0x90287e0,
>> socket_=0x902a210) at reaper.cpp:101
>> #7  0xf36fbfd4 in zmq::object_t::process_command (this=0x90287e0,
>> cmd_=...) at object.cpp:120
>> #8  0xf370591f in zmq::reaper_t::in_event (this=0x90287e0) at reaper.cpp:72
>> #9  0xf36f64f7 in zmq::epoll_t::loop (this=0x9029a10) at epoll.cpp:161
>> #10 0xf36f65c5 in zmq::epoll_t::worker_routine (arg_=0x9029a10) at
>> epoll.cpp:174 #11 0xf3710792 in thread_routine (arg_=0x9029a54) at
>> thread.cpp:75 #12 0xf7749adf in start_thread () from /lib/libpthread.so.0
>> #13 0x45e3655e in clone () from /lib/libc.so.6
>>
>> I can prevent the error by doing a zmq_term on the context or just
>> sleeping before returning control to the event loop but obviously
>> these solutions aren't optimal.
>>
>> Is there some way to block until zmq_close is really done?
>
> How about removing the socket's fd from the eventloop's monitoring at the time
> of close? It would be surprising if you can add an fd but not remove it...
>
> Justin
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From ivan.pechorin at gmail.com  Thu Jul  5 08:27:15 2012
From: ivan.pechorin at gmail.com (Ivan Pechorin)
Date: Thu, 5 Jul 2012 10:27:15 +0400
Subject: [zeromq-dev] How to safely use a socket in multiple threads in
 zmq 2.1.11
In-Reply-To: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
References: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
Message-ID: <CANmb91jFdJubz7sUnTxHVH7mMBA3PgzA5dk-vAm2KRzrO4dH7w@mail.gmail.com>

2012/7/5 Thiago Cangussu <cangussu at gmail.com>

>
> The question is: what should I do to make this multi-threaded usage of
> sockets safe? Is it enough to guarantee that multiple threads do not
> perform send/receive simultaneously?
>
>
This is already in the FAQ [1]:

"For those situations where a dedicated socket per thread is infeasible, a
socket may be shared *if and only if *each thread executes a full memory
barrier before accessing the socket. Most languages support a Mutex or
Spinlock which will execute the full memory barrier on your behalf."

In particular, for Linux/UNIX systems with pthreads, the memory barrier by
pthread_mutex_lock(), pthread_mutex_unlock() and a bunch of other pthread_*
functions is guaranteed by the POSIX specification [2].

So, if you don't have control on a number of threads that use ZeroMQ
sockets and the performance penalty of mutex lock/unlock on every operation
is not a problem for your needs, just make sure that your pool of sockets
is protected by mutexes.

[1] http://www.zeromq.org/area:faq
[2]
http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap04.html#tag_04_11
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/67e94ea6/attachment.htm 

From ivan.pechorin at gmail.com  Thu Jul  5 08:35:13 2012
From: ivan.pechorin at gmail.com (Ivan Pechorin)
Date: Thu, 5 Jul 2012 10:35:13 +0400
Subject: [zeromq-dev] How to safely use a socket in multiple threads in
 zmq 2.1.11
In-Reply-To: <20120704155715.63ffadfe@nehalam.linuxnetplumber.net>
References: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
	<20120704155715.63ffadfe@nehalam.linuxnetplumber.net>
Message-ID: <CANmb91hGj+_WkKHs4fKb=KMjWw1NujbyT6FZMrcNQvSO-a3Y7A@mail.gmail.com>

2012/7/5 Stephen Hemminger <shemminger at vyatta.com>

>
> You can have a per-thread pool of sockets, but you can't create socket
> in one thread and then use it in another.


You CAN create in one thread and then use it in another - just ensure
there's a memory barrier in the middle.


> Can you just have one socket
> per thread in thread local storage?
>

Sometime it's not feasible because you don't have control on these threads
and on their number.

For instance, if you develop some library or component that is intended to
run inside an existing application or app server, and this app or app
server uses lots (hundreds or even thousands) of threads, you will hit the
open file limit at some point if you create socket per thread when really
all you need is just a pool of few sockets.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/b0ffb641/attachment.htm 

From Sharon.Ben-Asher at avg.com  Thu Jul  5 08:51:58 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Thu, 5 Jul 2012 06:51:58 +0000
Subject: [zeromq-dev] Error: comparison between signed and unsigned integer
	expressions
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FBE75@ex10czdmb01>

Hello,

I am trying to build ZeroMQ build for Android.  I have a VM with Ubuntu 12.04 on x86_64, NDK is R8, configured as standalone toolchain, exactly like your websites instructs.
I downloaded the 3.x git project (git clone https://github.com/zeromq/zeromq3-x.git)
configure was ok.
When I do make, I get the following errors:

cc1plus: warnings being treated as errors
tcp_address.cpp: In constructor 'zmq::tcp_address_t::tcp_address_t(const sockaddr*, socklen_t)':
tcp_address.cpp:374: error: comparison between signed and unsigned integer expressions
tcp_address.cpp:377: error: comparison between signed and unsigned integer expressions
tcp_address.cpp: In member function 'const bool zmq::tcp_address_mask_t::match_address(const sockaddr*, socklen_t) const':
tcp_address.cpp:580: error: comparison between signed and unsigned integer expressions
make[2]: *** [libzmq_la-tcp_address.lo] Error 1
make[2]: Leaving directory `/home/sharon/zeromq3-x/src'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/sharon/zeromq3-x/src'
make: *** [all-recursive] Error 1

Attached is the full make output  + configure.log

Thanks,

Sharon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/bed79e59/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: make.out
Type: application/octet-stream
Size: 2667 bytes
Desc: make.out
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/bed79e59/attachment.obj 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: config.log
Type: application/octet-stream
Size: 88686 bytes
Desc: config.log
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/bed79e59/attachment-0001.obj 

From ph at imatix.com  Thu Jul  5 11:43:36 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Thu, 5 Jul 2012 18:43:36 +0900
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B2E7FBE75@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FBE75@ex10czdmb01>
Message-ID: <CADL5_siVXMH+qSDNtRz4DS9NOErYhsTF7vyHtPON9+-nWBcN8w@mail.gmail.com>

Hi Sharon,

Is it possible that socklen_t is signed on that compiler? Try
inserting a cast "(size_t) sa_len". If that works, we'll make a patch.

-Pieter

On Thu, Jul 5, 2012 at 3:51 PM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com> wrote:
> Hello,
>
>
>
> I am trying to build ZeroMQ build for Android.  I have a VM with Ubuntu
> 12.04 on x86_64, NDK is R8, configured as standalone toolchain, exactly like
> your websites instructs.
>
> I downloaded the 3.x git project (git clone
> https://github.com/zeromq/zeromq3-x.git)
>
> configure was ok.
>
> When I do make, I get the following errors:
>
>
>
> cc1plus: warnings being treated as errors
>
> tcp_address.cpp: In constructor 'zmq::tcp_address_t::tcp_address_t(const
> sockaddr*, socklen_t)':
>
> tcp_address.cpp:374: error: comparison between signed and unsigned integer
> expressions
>
> tcp_address.cpp:377: error: comparison between signed and unsigned integer
> expressions
>
> tcp_address.cpp: In member function 'const bool
> zmq::tcp_address_mask_t::match_address(const sockaddr*, socklen_t) const':
>
> tcp_address.cpp:580: error: comparison between signed and unsigned integer
> expressions
>
> make[2]: *** [libzmq_la-tcp_address.lo] Error 1
>
> make[2]: Leaving directory `/home/sharon/zeromq3-x/src'
>
> make[1]: *** [all] Error 2
>
> make[1]: Leaving directory `/home/sharon/zeromq3-x/src'
>
> make: *** [all-recursive] Error 1
>
>
>
> Attached is the full make output  + configure.log
>
>
>
> Thanks,
>
>
>
> Sharon
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From Sharon.Ben-Asher at avg.com  Thu Jul  5 11:56:26 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Thu, 5 Jul 2012 09:56:26 +0000
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FBECF@ex10czdmb01>

I solved the problem by converting the sizeof to int like this
if (sa->sa_family == AF_INET && sa_len >= (int)sizeof (address.ipv4)) {
I guess it's a platform specific issue? Anyway, it is needed to compile the lib for ARM.

From: Sharon Ben-Asher
Sent: Thursday, July 05, 2012 9:52 AM
To: 'zeromq-dev at lists.zeromq.org'
Subject: Error: comparison between signed and unsigned integer expressions

Hello,

I am trying to build ZeroMQ build for Android.  I have a VM with Ubuntu 12.04 on x86_64, NDK is R8, configured as standalone toolchain, exactly like your websites instructs.
I downloaded the 3.x git project (git clone https://github.com/zeromq/zeromq3-x.git)
configure was ok.
When I do make, I get the following errors:

cc1plus: warnings being treated as errors
tcp_address.cpp: In constructor 'zmq::tcp_address_t::tcp_address_t(const sockaddr*, socklen_t)':
tcp_address.cpp:374: error: comparison between signed and unsigned integer expressions
tcp_address.cpp:377: error: comparison between signed and unsigned integer expressions
tcp_address.cpp: In member function 'const bool zmq::tcp_address_mask_t::match_address(const sockaddr*, socklen_t) const':
tcp_address.cpp:580: error: comparison between signed and unsigned integer expressions
make[2]: *** [libzmq_la-tcp_address.lo] Error 1
make[2]: Leaving directory `/home/sharon/zeromq3-x/src'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/sharon/zeromq3-x/src'
make: *** [all-recursive] Error 1

Attached is the full make output  + configure.log

Thanks,

Sharon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/2091f86d/attachment.htm 

From ph at imatix.com  Thu Jul  5 12:02:33 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Thu, 5 Jul 2012 19:02:33 +0900
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B2E7FBECF@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FBECF@ex10czdmb01>
Message-ID: <CADL5_siDo9QAQOime0_1hXQvzu-=oeKMeo9ns42Mx9Sd7b3FMQ@mail.gmail.com>

Hi Sharon,

OK, so socklen_t is definitely signed on your platform. If you would
make an issue in Jira, I'll make a patch for all current 0MQ versions.

Thanks
Pieter

On Thu, Jul 5, 2012 at 6:56 PM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com> wrote:
> I solved the problem by converting the sizeof to int like this
>
> if (sa->sa_family == AF_INET && sa_len >= (int)sizeof (address.ipv4)) {
>
> I guess it?s a platform specific issue? Anyway, it is needed to compile the
> lib for ARM.
>
>
>
> From: Sharon Ben-Asher
> Sent: Thursday, July 05, 2012 9:52 AM
> To: 'zeromq-dev at lists.zeromq.org'
> Subject: Error: comparison between signed and unsigned integer expressions
>
>
>
> Hello,
>
>
>
> I am trying to build ZeroMQ build for Android.  I have a VM with Ubuntu
> 12.04 on x86_64, NDK is R8, configured as standalone toolchain, exactly like
> your websites instructs.
>
> I downloaded the 3.x git project (git clone
> https://github.com/zeromq/zeromq3-x.git)
>
> configure was ok.
>
> When I do make, I get the following errors:
>
>
>
> cc1plus: warnings being treated as errors
>
> tcp_address.cpp: In constructor 'zmq::tcp_address_t::tcp_address_t(const
> sockaddr*, socklen_t)':
>
> tcp_address.cpp:374: error: comparison between signed and unsigned integer
> expressions
>
> tcp_address.cpp:377: error: comparison between signed and unsigned integer
> expressions
>
> tcp_address.cpp: In member function 'const bool
> zmq::tcp_address_mask_t::match_address(const sockaddr*, socklen_t) const':
>
> tcp_address.cpp:580: error: comparison between signed and unsigned integer
> expressions
>
> make[2]: *** [libzmq_la-tcp_address.lo] Error 1
>
> make[2]: Leaving directory `/home/sharon/zeromq3-x/src'
>
> make[1]: *** [all] Error 2
>
> make[1]: Leaving directory `/home/sharon/zeromq3-x/src'
>
> make: *** [all-recursive] Error 1
>
>
>
> Attached is the full make output  + configure.log
>
>
>
> Thanks,
>
>
>
> Sharon
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From ivan.pechorin at gmail.com  Thu Jul  5 12:09:14 2012
From: ivan.pechorin at gmail.com (Ivan Pechorin)
Date: Thu, 5 Jul 2012 14:09:14 +0400
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
In-Reply-To: <CADL5_siDo9QAQOime0_1hXQvzu-=oeKMeo9ns42Mx9Sd7b3FMQ@mail.gmail.com>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FBECF@ex10czdmb01>
	<CADL5_siDo9QAQOime0_1hXQvzu-=oeKMeo9ns42Mx9Sd7b3FMQ@mail.gmail.com>
Message-ID: <CANmb91ho7Cq_7NfEP65ZT_CuSM9CWq+9qg+xXEvRTbTsqVODDg@mail.gmail.com>

2012/7/5 Pieter Hintjens <ph at imatix.com>

> Hi Sharon,
>
> OK, so socklen_t is definitely signed on your platform. If you would
> make an issue in Jira, I'll make a patch for all current 0MQ versions.
>
>
Android NDK has the following typedef:

typedef int socklen_t

I wonder why they choose to go against UNIX specs that require socklen_t to
be unsigned (see
http://pubs.opengroup.org/onlinepubs/7908799/xns/syssocket.h.html)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/6784fcb1/attachment.htm 

From Sharon.Ben-Asher at avg.com  Thu Jul  5 12:28:32 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Thu, 5 Jul 2012 10:28:32 +0000
Subject: [zeromq-dev] Error: cannot link with -lzmq
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FBF01@ex10czdmb01>

Hello again,

After successful build of the C lib, I have a problem with building zmq.jar:

configure:15622: checking for zmq_init in -lzmq
configure:15647: g++ -o conftest -g -O2 -D_REENTRANT -D_THREAD_SAFE -fPIC -I/home/sharon/zeromq-android/include  -L/home/sharon/zeromq-android/lib conftest.cpp -lzmq   >&5
/usr/bin/ld: skipping incompatible /home/sharon/zeromq-android/lib/libzmq.so when searching for -lzmq
/usr/bin/ld: skipping incompatible /home/sharon/zeromq-android/lib/libzmq.a when searching for -lzmq
/usr/bin/ld: cannot find -lzmq
collect2: ld returned 1 exit status

Thanks

Sharon


From: Sharon Ben-Asher
Sent: Thursday, July 05, 2012 12:56 PM
To: 'zeromq-dev at lists.zeromq.org'
Subject: RE: Error: comparison between signed and unsigned integer expressions

I solved the problem by converting the sizeof to int like this
if (sa->sa_family == AF_INET && sa_len >= (int)sizeof (address.ipv4)) {
I guess it's a platform specific issue? Anyway, it is needed to compile the lib for ARM.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/d1bd988f/attachment.htm 

From bb.varga at gmail.com  Thu Jul  5 12:34:48 2012
From: bb.varga at gmail.com (=?UTF-8?Q?Bal=C3=A1zs_Varga?=)
Date: Thu, 5 Jul 2012 12:34:48 +0200
Subject: [zeromq-dev] Delphi binding.
Message-ID: <CAFaqPK+7V6KnCGX8XecGTD7vVOZsMP=ik6J2EsJD95VZ8uKCDQ@mail.gmail.com>

Dear members,

I just want to let you know, that I've started working on a 0MQ binding for
Delphi, and pushed the initial version of the code, which is here:
https://github.com/bvarga/delphizmq . For now it's not well tested, however
20 examples translated from the guide are working great for me.

bye
balazs.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/bf82194f/attachment.htm 

From cangussu at gmail.com  Thu Jul  5 14:13:29 2012
From: cangussu at gmail.com (Thiago Cangussu)
Date: Thu, 5 Jul 2012 09:13:29 -0300
Subject: [zeromq-dev] How to safely use a socket in multiple threads in
 zmq 2.1.11
In-Reply-To: <CANmb91jFdJubz7sUnTxHVH7mMBA3PgzA5dk-vAm2KRzrO4dH7w@mail.gmail.com>
References: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
	<CANmb91jFdJubz7sUnTxHVH7mMBA3PgzA5dk-vAm2KRzrO4dH7w@mail.gmail.com>
Message-ID: <CANP5a=V2tj870Lfp2gSa08UKc5pu_5SnWYMrPrGqN=rTVUbNOg@mail.gmail.com>

On Thu, Jul 5, 2012 at 3:27 AM, Ivan Pechorin <ivan.pechorin at gmail.com>wrote:

>
>
> 2012/7/5 Thiago Cangussu <cangussu at gmail.com>
>
>>
>> The question is: what should I do to make this multi-threaded usage of
>> sockets safe? Is it enough to guarantee that multiple threads do not
>> perform send/receive simultaneously?
>>
>>
> This is already in the FAQ [1]:
>
> "For those situations where a dedicated socket per thread is infeasible, a
> socket may be shared *if and only if *each thread executes a full memory
> barrier before accessing the socket. Most languages support a Mutex or
> Spinlock which will execute the full memory barrier on your behalf."
>

My bad, I only checked the guide/man-pages and was a little unsure that a
mutex would do the trick.



>  In particular, for Linux/UNIX systems with pthreads, the memory barrier
> by pthread_mutex_lock(), pthread_mutex_unlock() and a bunch of other
> pthread_* functions is guaranteed by the POSIX specification [2].
>
> So, if you don't have control on a number of threads that use ZeroMQ
> sockets and the performance penalty of mutex lock/unlock on every operation
> is not a problem for your needs, just make sure that your pool of sockets
> is protected by mutexes.
>

That's the case, I was ok with the performance penalty of creating/closing
the socket, but the file descriptor limit is a problem. The mutex
lock/unlock will not be an issue.

Thanks a lot,
Thiago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/e03d9c1b/attachment.htm 

From cangussu at gmail.com  Thu Jul  5 14:19:57 2012
From: cangussu at gmail.com (Thiago Cangussu)
Date: Thu, 5 Jul 2012 09:19:57 -0300
Subject: [zeromq-dev] How to safely use a socket in multiple threads in
 zmq 2.1.11
In-Reply-To: <CANmb91hGj+_WkKHs4fKb=KMjWw1NujbyT6FZMrcNQvSO-a3Y7A@mail.gmail.com>
References: <CANP5a=VJ6Ah4QMJaVfQOJW2tY_wk9Q7jy_pkSStc1nW+vv-mHg@mail.gmail.com>
	<20120704155715.63ffadfe@nehalam.linuxnetplumber.net>
	<CANmb91hGj+_WkKHs4fKb=KMjWw1NujbyT6FZMrcNQvSO-a3Y7A@mail.gmail.com>
Message-ID: <CANP5a=XAbEwxCn2ggFkcSyrTr=cxQ6DRRb=jFQngq8UQrBdAkw@mail.gmail.com>

On Thu, Jul 5, 2012 at 3:35 AM, Ivan Pechorin <ivan.pechorin at gmail.com>wrote:

> Can you just have one socket
>> per thread in thread local storage?
>>
>
> Sometime it's not feasible because you don't have control on these threads
> and on their number.
>
> For instance, if you develop some library or component that is intended to
> run inside an existing application or app server, and this app or app
> server uses lots (hundreds or even thousands) of threads, you will hit the
> open file limit at some point if you create socket per thread when really
> all you need is just a pool of few sockets.
>

That's exactly the case. And there is also the problem that I need to close
the sockets before calling zmq_term. I actually tried this solution (TLS)
but it started to look like a code smell, so I revisited the muti-thread
use of sockets.

As a side note, the guide really did a good job keeping me away from using
sockets from multiple threads :)

Thanks,
Thiago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/27c0ae4e/attachment.htm 

From steven.mccoy at miru.hk  Thu Jul  5 15:12:02 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Thu, 5 Jul 2012 09:12:02 -0400
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
In-Reply-To: <CANmb91ho7Cq_7NfEP65ZT_CuSM9CWq+9qg+xXEvRTbTsqVODDg@mail.gmail.com>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FBECF@ex10czdmb01>
	<CADL5_siDo9QAQOime0_1hXQvzu-=oeKMeo9ns42Mx9Sd7b3FMQ@mail.gmail.com>
	<CANmb91ho7Cq_7NfEP65ZT_CuSM9CWq+9qg+xXEvRTbTsqVODDg@mail.gmail.com>
Message-ID: <CAGFWLiiMKgpwbJiS5nYPKm82RtWi7Q4HObKzQ5B7=B=qjwTDgQ@mail.gmail.com>

On 5 July 2012 06:09, Ivan Pechorin <ivan.pechorin at gmail.com> wrote:

> Android NDK has the following typedef:
>
> typedef int socklen_t
>
> I wonder why they choose to go against UNIX specs that require socklen_t
> to be unsigned (see
> http://pubs.opengroup.org/onlinepubs/7908799/xns/syssocket.h.html)
>

It appears to be a BSD trait,here's a rather moody article about it:

http://everything2.com/title/socklen_t

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/9de6415e/attachment.htm 

From diffuser78 at gmail.com  Thu Jul  5 17:23:53 2012
From: diffuser78 at gmail.com (Diffuser78)
Date: Thu, 5 Jul 2012 08:23:53 -0700
Subject: [zeromq-dev] Design Suggestion for simple app using zmq
Message-ID: <CAKni0ecimZWt=YxKpcYoBFWLo=_ZOwkTWOWQmmHtpN8UR65qQA@mail.gmail.com>

I have a client and a server app. Client needs to send 1000 lightweight
(5KBs) messages per second to the server. Client also wants an ACK from
server that server got the msg that client sent.

Here is what I have designed. Please let me know your thought in case you
see incorrect usage of the patterns.


CLIENT (*DEALER* socket)

Do the following in while loop

1. Get messages to send
      - Send then using zmq_send()
2. zmq_poll() for fds ready for reading
     - Use zmq_recv() to receive the messages


SERVER (*ROUTER* socket)

Do the following in while loop

1. zmq_poll () for fds ready for reading
    - Use zmq_recv() to receive the messages
2. Perform Processing (Business Logic)
3. Get messages to send
      - Send then using zmq_send()


Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/1bed8a32/attachment.htm 

From paul at colomiets.name  Thu Jul  5 22:02:40 2012
From: paul at colomiets.name (Paul Colomiets)
Date: Thu, 5 Jul 2012 23:02:40 +0300
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>
Message-ID: <CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>

Hi Mark, Ian,

On Thu, Jul 5, 2012 at 12:44 AM, Ian Barber <ian.barber at gmail.com> wrote:
>> I?m evaluating zeroMQ, and one main criteria we have is to support
>> Websockets for client to server communications.
>>
>>
>>
>> Has there been any work on making a native Websocket transport for ZeroMQ
>> in the core C++ libraries ?
>
>
> No work that made it anywhere near head, but as Apostolis said there are
> plenty of bridging methods. My personal favourite is Paul's ZeroGW:
> https://github.com/tailhook/zerogw
>

Thanks Ian, I appreciate your respect :) However, zerogw is designed
with browser clients in mind. Mark's use case may be a bit different.
Although, I would support usage of zerogw for other applications, but
I'm not going to implement outbound websocket connections at the
current stage of evolution of zerogw, so you need some complementary
client implementation.

-- 
Paul

From mark.farnan at petrolink.com  Fri Jul  6 00:53:18 2012
From: mark.farnan at petrolink.com (Mark Farnan)
Date: Thu, 5 Jul 2012 17:53:18 -0500
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>
	<CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>
Message-ID: <043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>

Correct,  What I am after is a bit different.

I am looking for using Websockets as native transport,  machine to machine
running ZeroMQ natively at each end, across firewalls.   For our case we
can't rely on a TCP port being open to work through.  Websockets provides
the ideal transport protocol for this.

i.e. 
 Connects Https
 Handles Security (even  basic Auth is fine, as it is HttpS).
 Upgrades the connection to Websocket  wss
 ZeroMQ runs 'as is' over the now established Websocket.

This, or something similar, is what I am thinking.   (You can actually do
all of this in one step with a http connect and upgrade frame)

So ideally a client would provide the address,  i.e.
"wss://someserver.someplace.com/something", to the Socket Connect function,
along with an optional username and PW,      It establishes the session, and
ZeroMQ works as before from there.

The ZeroMQ receiving the connection would need a registered callback of some
kind to handle  the authentication, which just passes back 'true or false'
if its accepted or not. 


Would anyone else maybe interested in doing this ?

Regards

Mark

-----Original Message-----
From: zeromq-dev-bounces at lists.zeromq.org
[mailto:zeromq-dev-bounces at lists.zeromq.org] On Behalf Of Paul Colomiets
Sent: Thursday, July 05, 2012 3:03 PM
To: ZeroMQ development list
Subject: Re: [zeromq-dev] Websockets as a Transport ?

Hi Mark, Ian,

On Thu, Jul 5, 2012 at 12:44 AM, Ian Barber <ian.barber at gmail.com> wrote:
>> I'm evaluating zeroMQ, and one main criteria we have is to support 
>> Websockets for client to server communications.
>>
>>
>>
>> Has there been any work on making a native Websocket transport for 
>> ZeroMQ in the core C++ libraries ?
>
>
> No work that made it anywhere near head, but as Apostolis said there 
> are plenty of bridging methods. My personal favourite is Paul's ZeroGW:
> https://github.com/tailhook/zerogw
>

Thanks Ian, I appreciate your respect :) However, zerogw is designed with
browser clients in mind. Mark's use case may be a bit different.
Although, I would support usage of zerogw for other applications, but I'm
not going to implement outbound websocket connections at the current stage
of evolution of zerogw, so you need some complementary client
implementation.

--
Paul
_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From ian.barber at gmail.com  Fri Jul  6 01:27:24 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Fri, 6 Jul 2012 00:27:24 +0100
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>
	<CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>
	<043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
Message-ID: <CAEYnEwFonEmjz8XPYOrQ-yTvPGCwqr=nU7QZxEA4_V46+fnRLw@mail.gmail.com>

On Thu, Jul 5, 2012 at 11:53 PM, Mark Farnan <mark.farnan at petrolink.com>wrote:

> Correct,  What I am after is a bit different.
>
> I am looking for using Websockets as native transport,  machine to machine
> running ZeroMQ natively at each end, across firewalls.   For our case we
> can't rely on a TCP port being open to work through.  Websockets provides
> the ideal transport protocol for this.
>
> i.e.
>  Connects Https
>  Handles Security (even  basic Auth is fine, as it is HttpS).
>  Upgrades the connection to Websocket  wss
>  ZeroMQ runs 'as is' over the now established Websocket.


It's a non-trivial bit of work to implement a new transport, but it's
definitely an option. I'll doubt you'll get someone willing to jump in and
do it unless someone is quietly already working on it though, as it'd be a
fair investment of time. If you're interested in doing it yourself though
people here can certainly point you in the right direction.

Other options worth considering would be:

1 - Pieter's VTX: https://github.com/imatix/vtx - this is designed to be an
easier way of adding extra protocols, building on top of the existing
zeromq library.
2 - Running ZeroMQ TCP across port 80 or 443. Depending on the firewall
config, you might find this is enough to get the traffic through and will
get you up and running with a lot less effort. YMMV.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120706/6c54393d/attachment.htm 

From mark.farnan at petrolink.com  Fri Jul  6 01:39:01 2012
From: mark.farnan at petrolink.com (Mark Farnan)
Date: Thu, 5 Jul 2012 18:39:01 -0500
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <CAEYnEwFonEmjz8XPYOrQ-yTvPGCwqr=nU7QZxEA4_V46+fnRLw@mail.gmail.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>	<CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>	<043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
	<CAEYnEwFonEmjz8XPYOrQ-yTvPGCwqr=nU7QZxEA4_V46+fnRLw@mail.gmail.com>
Message-ID: <005e01cd5b07$5b4ddaa0$11e98fe0$@petrolink.com>

Thanks for the quick reply.

 

I'll take a look at VTX.

 

Sadly, we already know that using ZeroMQ on Port 80/443 won't fly at all.
Been down this road already with trying to use a TCP socket on these ports.


Regards

 

Mark Farnan

Director, Product Development

Petrolink Services Inc

+1 281 908 9814

 

From: zeromq-dev-bounces at lists.zeromq.org
[mailto:zeromq-dev-bounces at lists.zeromq.org] On Behalf Of Ian Barber
Sent: Thursday, July 05, 2012 6:27 PM
To: ZeroMQ development list
Subject: Re: [zeromq-dev] Websockets as a Transport ?

 

 

On Thu, Jul 5, 2012 at 11:53 PM, Mark Farnan <mark.farnan at petrolink.com>
wrote:

Correct,  What I am after is a bit different.

I am looking for using Websockets as native transport,  machine to machine
running ZeroMQ natively at each end, across firewalls.   For our case we
can't rely on a TCP port being open to work through.  Websockets provides
the ideal transport protocol for this.

i.e.
 Connects Https
 Handles Security (even  basic Auth is fine, as it is HttpS).
 Upgrades the connection to Websocket  wss
 ZeroMQ runs 'as is' over the now established Websocket.

 

It's a non-trivial bit of work to implement a new transport, but it's
definitely an option. I'll doubt you'll get someone willing to jump in and
do it unless someone is quietly already working on it though, as it'd be a
fair investment of time. If you're interested in doing it yourself though
people here can certainly point you in the right direction. 

 

Other options worth considering would be:

 

1 - Pieter's VTX: https://github.com/imatix/vtx - this is designed to be an
easier way of adding extra protocols, building on top of the existing zeromq
library. 

2 - Running ZeroMQ TCP across port 80 or 443. Depending on the firewall
config, you might find this is enough to get the traffic through and will
get you up and running with a lot less effort. YMMV. 

 

Ian

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/deebe02b/attachment.htm 

From diffuser78 at gmail.com  Fri Jul  6 01:58:53 2012
From: diffuser78 at gmail.com (Diffuser78)
Date: Thu, 5 Jul 2012 16:58:53 -0700
Subject: [zeromq-dev] Design Suggestion for simple app using zmq
In-Reply-To: <CAKni0ecimZWt=YxKpcYoBFWLo=_ZOwkTWOWQmmHtpN8UR65qQA@mail.gmail.com>
References: <CAKni0ecimZWt=YxKpcYoBFWLo=_ZOwkTWOWQmmHtpN8UR65qQA@mail.gmail.com>
Message-ID: <CAKni0eczsTTDkYtUXO=5TRbKazQVH4=aisrHS9-64YTG+SpYog@mail.gmail.com>

Anyone ?

On Thu, Jul 5, 2012 at 8:23 AM, Diffuser78 <diffuser78 at gmail.com> wrote:

> I have a client and a server app. Client needs to send 1000 lightweight
> (5KBs) messages per second to the server. Client also wants an ACK from
> server that server got the msg that client sent.
>
> Here is what I have designed. Please let me know your thought in case you
> see incorrect usage of the patterns.
>
>
> CLIENT (*DEALER* socket)
>
> Do the following in while loop
>
> 1. Get messages to send
>       - Send then using zmq_send()
> 2. zmq_poll() for fds ready for reading
>      - Use zmq_recv() to receive the messages
>
>
> SERVER (*ROUTER* socket)
>
> Do the following in while loop
>
> 1. zmq_poll () for fds ready for reading
>     - Use zmq_recv() to receive the messages
> 2. Perform Processing (Business Logic)
> 3. Get messages to send
>       - Send then using zmq_send()
>
>
> Thanks.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/dbea23b7/attachment.htm 

From jhawk28 at gmail.com  Fri Jul  6 04:52:29 2012
From: jhawk28 at gmail.com (Joshua Foster)
Date: Thu, 5 Jul 2012 22:52:29 -0400
Subject: [zeromq-dev] Design Suggestion for simple app using zmq
In-Reply-To: <CAKni0eczsTTDkYtUXO=5TRbKazQVH4=aisrHS9-64YTG+SpYog@mail.gmail.com>
References: <CAKni0ecimZWt=YxKpcYoBFWLo=_ZOwkTWOWQmmHtpN8UR65qQA@mail.gmail.com>
	<CAKni0eczsTTDkYtUXO=5TRbKazQVH4=aisrHS9-64YTG+SpYog@mail.gmail.com>
Message-ID: <9C9647F5-F8B5-4A52-913A-5EEE441D68EC@gmail.com>

Are the ACK's async? If they are not, you can use the REQ socket instead. Are you using multiple threads? 

Joshua

On Jul 5, 2012, at 7:58 PM, Diffuser78 wrote:

> Anyone ?
> 
> On Thu, Jul 5, 2012 at 8:23 AM, Diffuser78 <diffuser78 at gmail.com> wrote:
> I have a client and a server app. Client needs to send 1000 lightweight (5KBs) messages per second to the server. Client also wants an ACK from server that server got the msg that client sent. 
> 
> Here is what I have designed. Please let me know your thought in case you see incorrect usage of the patterns.
> 
> 
> CLIENT (DEALER socket)
> 
> Do the following in while loop
> 
> 1. Get messages to send
>       - Send then using zmq_send()
> 2. zmq_poll() for fds ready for reading
>      - Use zmq_recv() to receive the messages
>      
> 
> SERVER (ROUTER socket)
> 
> Do the following in while loop
> 
> 1. zmq_poll () for fds ready for reading
>     - Use zmq_recv() to receive the messages
> 2. Perform Processing (Business Logic)
> 3. Get messages to send
>       - Send then using zmq_send()
>        
> 
> Thanks.
> 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120705/5f828ab7/attachment.htm 

From ph at imatix.com  Fri Jul  6 05:27:34 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 6 Jul 2012 12:27:34 +0900
Subject: [zeromq-dev] Delphi binding.
In-Reply-To: <CAFaqPK+7V6KnCGX8XecGTD7vVOZsMP=ik6J2EsJD95VZ8uKCDQ@mail.gmail.com>
References: <CAFaqPK+7V6KnCGX8XecGTD7vVOZsMP=ik6J2EsJD95VZ8uKCDQ@mail.gmail.com>
Message-ID: <CADL5_siV2OUTUm9XjthWdb_XyPD0pQHKd1EHf3YNq5nUW2_JrA@mail.gmail.com>

Nice... :)

On Thu, Jul 5, 2012 at 7:34 PM, Bal?zs Varga <bb.varga at gmail.com> wrote:
> Dear members,
>
> I just want to let you know, that I've started working on a 0MQ binding for
> Delphi, and pushed the initial version of the code, which is here:
> https://github.com/bvarga/delphizmq . For now it's not well tested, however
> 20 examples translated from the guide are working great for me.
>
> bye
> balazs.
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From hurtonm at gmail.com  Fri Jul  6 10:28:32 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Fri, 6 Jul 2012 10:28:32 +0200
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B2E7FBE75@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FBE75@ex10czdmb01>
Message-ID: <CALdZE=t4a_p7KFTTRs7acFpoq7RjxSSaz2yO5R_EEJ3UU_aj-Q@mail.gmail.com>

Hi Sharon, what revision are you using?
Commit eb6c668c2ae961a4a53d11d81dd6d2df5bffa666 from Javier Lopez
already fixes that.

- Martin

On Thu, Jul 5, 2012 at 8:51 AM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com> wrote:
> Hello,
>
>
>
> I am trying to build ZeroMQ build for Android.  I have a VM with Ubuntu
> 12.04 on x86_64, NDK is R8, configured as standalone toolchain, exactly like
> your websites instructs.
>
> I downloaded the 3.x git project (git clone
> https://github.com/zeromq/zeromq3-x.git)
>
> configure was ok.
>
> When I do make, I get the following errors:
>
>
>
> cc1plus: warnings being treated as errors
>
> tcp_address.cpp: In constructor 'zmq::tcp_address_t::tcp_address_t(const
> sockaddr*, socklen_t)':
>
> tcp_address.cpp:374: error: comparison between signed and unsigned integer
> expressions
>
> tcp_address.cpp:377: error: comparison between signed and unsigned integer
> expressions
>
> tcp_address.cpp: In member function 'const bool
> zmq::tcp_address_mask_t::match_address(const sockaddr*, socklen_t) const':
>
> tcp_address.cpp:580: error: comparison between signed and unsigned integer
> expressions
>
> make[2]: *** [libzmq_la-tcp_address.lo] Error 1
>
> make[2]: Leaving directory `/home/sharon/zeromq3-x/src'
>
> make[1]: *** [all] Error 2
>
> make[1]: Leaving directory `/home/sharon/zeromq3-x/src'
>
> make: *** [all-recursive] Error 1
>
>
>
> Attached is the full make output  + configure.log
>
>
>
> Thanks,
>
>
>
> Sharon
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From ph at imatix.com  Fri Jul  6 10:34:27 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 6 Jul 2012 17:34:27 +0900
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
In-Reply-To: <CALdZE=t4a_p7KFTTRs7acFpoq7RjxSSaz2yO5R_EEJ3UU_aj-Q@mail.gmail.com>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FBE75@ex10czdmb01>
	<CALdZE=t4a_p7KFTTRs7acFpoq7RjxSSaz2yO5R_EEJ3UU_aj-Q@mail.gmail.com>
Message-ID: <CADL5_sgj131tsYP_74fxN3=QjzF3FFiRdrXOBFic8p=FNK=Ekg@mail.gmail.com>

On Fri, Jul 6, 2012 at 5:28 PM, Martin Hurton <hurtonm at gmail.com> wrote:
> Hi Sharon, what revision are you using?
> Commit eb6c668c2ae961a4a53d11d81dd6d2df5bffa666 from Javier Lopez
> already fixes that.

:-) Lol, and my patch unfixes it. Will revert.

-Pieter

From rplehmann at gmail.com  Fri Jul  6 10:39:06 2012
From: rplehmann at gmail.com (Rene-Pierre Lehmann)
Date: Fri, 06 Jul 2012 19:39:06 +1100
Subject: [zeromq-dev] Full-duplex tcp sockets on io threads
Message-ID: <3bc319aab14294a9be67ee62dfc789f3@ripi.net>

Hi,

Is it possible with zeromq to use underlying tcp sockets in a 
full-duplex fashion?
By full duplex I mean that an io thread may read from a tcp socket 
while another io thread may concurrently write to the same socket.

Cheers, Rene-Pierre.

From ph at imatix.com  Fri Jul  6 12:57:59 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 6 Jul 2012 19:57:59 +0900
Subject: [zeromq-dev] Full-duplex tcp sockets on io threads
In-Reply-To: <3bc319aab14294a9be67ee62dfc789f3@ripi.net>
References: <3bc319aab14294a9be67ee62dfc789f3@ripi.net>
Message-ID: <CADL5_shULrk457byjqA8gSkDDCmZwDO1bv8rAir=6PZvpd0_AQ@mail.gmail.com>

On Fri, Jul 6, 2012 at 5:39 PM, Rene-Pierre Lehmann <rplehmann at gmail.com> wrote:

> Is it possible with zeromq to use underlying tcp sockets in a
> full-duplex fashion?
> By full duplex I mean that an io thread may read from a tcp socket
> while another io thread may concurrently write to the same socket.

No, this will cause the library to crash. However you can create two
sockets in two threads, and use one for input, and one for output.

-Pieter

From mek at mek.uz.ua  Fri Jul  6 16:08:48 2012
From: mek at mek.uz.ua (Max Kuznecov)
Date: Fri, 6 Jul 2012 17:08:48 +0300
Subject: [zeromq-dev] Using zeroMQ socket from freeradius module
Message-ID: <CAGb0cFmw8KhrsLXGvO+4KZdcEeP9ffhSJD_QvZdmj18EeJXH3w@mail.gmail.com>

Hi!

I'm facing a strange problem which I'm unable to resolve for about a
week, so I decided to ask in the list as a last resort.

I've got a system which collects requests from custom freeradius
account module, stores them in persistent queue on disk and then
serves to the agents, which put them into postgres afterwards.
Here's a brief arch image: http://imageshack.us/photo/my-images/821/archn.png/

In radius module a single 0mq socket (REQ) is instantiated which is
later used for sending requests to collector.

Collector is a separate process which binds on the REP socket (tcp, on
the same port as radius module, of course)
Collector's loop code: http://pastebin.com/xuN4pitv

Radius module code: http://pastebin.com/HhUK4nGG
It uses mutexes as radius calls the function from various threads.

xsend function is a basic wrapper: http://pastebin.com/V3dG0kV9

So, the problem is: when the radius initiates the module, it
successfully creates the context and the socket.
But, when it tries to send a message to the collector, the following happens:
Send operation reports ok (code 0), but when it tries to receive
afterwards it blocks. And the collector doesn't receive the data sent
by module!

I've put a lot of debug printfs all around and the collector just
doesn't seem to complete the recv function call.
But what is the most strange in all this is that if I run radius with
debug flag (-X) which tell it to run in single threaded and in
foreground - everything works just as planned - modules sends messages
to collector, collector replies etc. Even when I supply a -f option to
radiusd, which just evades calling fork() it still works. It doesn't
only when radius daemon is run in background.

So, basically this is what I cannot understand, if it didn't work in
any case I'd understand, but this...
Maybe anybody could came up with any idea about what's wrong here. I'd
be greatly appreciated!

-- 
~syhpoon

From mek at mek.uz.ua  Fri Jul  6 16:11:20 2012
From: mek at mek.uz.ua (Max Kuznecov)
Date: Fri, 6 Jul 2012 17:11:20 +0300
Subject: [zeromq-dev] Using zeroMQ socket from freeradius module
In-Reply-To: <CAGb0cFmw8KhrsLXGvO+4KZdcEeP9ffhSJD_QvZdmj18EeJXH3w@mail.gmail.com>
References: <CAGb0cFmw8KhrsLXGvO+4KZdcEeP9ffhSJD_QvZdmj18EeJXH3w@mail.gmail.com>
Message-ID: <CAGb0cF=DHR3muZ1aDUhmc_FA6bOd3r82Q7r3tezOK7JZd9v7Yw@mail.gmail.com>

Oh, sorry I forgot to mention  versions:
linux centos 6.2, zeromq 2.2.0

2012/7/6 Max Kuznecov <mek at mek.uz.ua>:
> Hi!
>
> I'm facing a strange problem which I'm unable to resolve for about a
> week, so I decided to ask in the list as a last resort.
>
> I've got a system which collects requests from custom freeradius
> account module, stores them in persistent queue on disk and then
> serves to the agents, which put them into postgres afterwards.
> Here's a brief arch image: http://imageshack.us/photo/my-images/821/archn.png/
>
> In radius module a single 0mq socket (REQ) is instantiated which is
> later used for sending requests to collector.
>
> Collector is a separate process which binds on the REP socket (tcp, on
> the same port as radius module, of course)
> Collector's loop code: http://pastebin.com/xuN4pitv
>
> Radius module code: http://pastebin.com/HhUK4nGG
> It uses mutexes as radius calls the function from various threads.
>
> xsend function is a basic wrapper: http://pastebin.com/V3dG0kV9
>
> So, the problem is: when the radius initiates the module, it
> successfully creates the context and the socket.
> But, when it tries to send a message to the collector, the following happens:
> Send operation reports ok (code 0), but when it tries to receive
> afterwards it blocks. And the collector doesn't receive the data sent
> by module!
>
> I've put a lot of debug printfs all around and the collector just
> doesn't seem to complete the recv function call.
> But what is the most strange in all this is that if I run radius with
> debug flag (-X) which tell it to run in single threaded and in
> foreground - everything works just as planned - modules sends messages
> to collector, collector replies etc. Even when I supply a -f option to
> radiusd, which just evades calling fork() it still works. It doesn't
> only when radius daemon is run in background.
>
> So, basically this is what I cannot understand, if it didn't work in
> any case I'd understand, but this...
> Maybe anybody could came up with any idea about what's wrong here. I'd
> be greatly appreciated!
>
> --
> ~syhpoon



-- 
~syhpoon

From diffuser78 at gmail.com  Fri Jul  6 17:08:24 2012
From: diffuser78 at gmail.com (Diffuser78)
Date: Fri, 6 Jul 2012 08:08:24 -0700
Subject: [zeromq-dev] Design Suggestion for simple app using zmq
In-Reply-To: <9C9647F5-F8B5-4A52-913A-5EEE441D68EC@gmail.com>
References: <CAKni0ecimZWt=YxKpcYoBFWLo=_ZOwkTWOWQmmHtpN8UR65qQA@mail.gmail.com>
	<CAKni0eczsTTDkYtUXO=5TRbKazQVH4=aisrHS9-64YTG+SpYog@mail.gmail.com>
	<9C9647F5-F8B5-4A52-913A-5EEE441D68EC@gmail.com>
Message-ID: <CAKni0ec-5DOAK1qZU4rKooWzYwSLz9dB4uZJQxt=KwsoGDyvTA@mail.gmail.com>

There is no particular requirement for the ACKs to be async. The only
requirement is that I should be able to may server's ACK to the message
that client had sent ? Is there a deterministic way to do that ?

I am not planning to use multiple threads.

Any other thoughts are welcome.

Many thanks.

On Thu, Jul 5, 2012 at 7:52 PM, Joshua Foster <jhawk28 at gmail.com> wrote:

> Are the ACK's async? If they are not, you can use the REQ socket instead.
> Are you using multiple threads?
>
> Joshua
>
> On Jul 5, 2012, at 7:58 PM, Diffuser78 wrote:
>
> Anyone ?
>
> On Thu, Jul 5, 2012 at 8:23 AM, Diffuser78 <diffuser78 at gmail.com> wrote:
>
>> I have a client and a server app. Client needs to send 1000 lightweight
>> (5KBs) messages per second to the server. Client also wants an ACK from
>> server that server got the msg that client sent.
>>
>> Here is what I have designed. Please let me know your thought in case you
>> see incorrect usage of the patterns.
>>
>>
>> CLIENT (*DEALER* socket)
>>
>> Do the following in while loop
>>
>> 1. Get messages to send
>>       - Send then using zmq_send()
>> 2. zmq_poll() for fds ready for reading
>>      - Use zmq_recv() to receive the messages
>>
>>
>> SERVER (*ROUTER* socket)
>>
>> Do the following in while loop
>>
>> 1. zmq_poll () for fds ready for reading
>>     - Use zmq_recv() to receive the messages
>> 2. Perform Processing (Business Logic)
>> 3. Get messages to send
>>       - Send then using zmq_send()
>>
>>
>> Thanks.
>>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120706/bee3cbad/attachment.htm 

From bklooste at gmail.com  Fri Jul  6 17:41:04 2012
From: bklooste at gmail.com (Bennie Kloosteman)
Date: Fri, 6 Jul 2012 23:41:04 +0800
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>
	<CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>
	<043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
Message-ID: <CAJT18ibGppGntbXcvtE=vpPd52nHmfaPiikhAZ+Ky7hRq+ZiYA@mail.gmail.com>

Hi Mark,

Web socket uses tcp underneath so getting through the firewall is no issue
just use port 80  , and in fact web sockets ( and things that use port 80 )
have issues with some ISP html proxy servers and load balancers and it will
take time for these products to mature to handle web sockets.

AFAIK web sockets is also client to server and most clients already have
open all ports out and few in , if you want in a solution like Microsoft
Azure uses is ideal ( they have a tcp connection , it tries to open a
firewall port to receive and if it fails it relays it via the cloud - but
in the best case you have direct client to client connections)

 I would not trust https too much it does not add a lot , make sure the
authentication is good.

Also  by the time you add the higher level costs like encryption what is
the point to using a light weight protocol  like zeromq ? You may as well
use a nice API like WS-Eventing  which can still do 3ms  packets  (
100Kpackets/sec per server)

Ben

On Fri, Jul 6, 2012 at 6:53 AM, Mark Farnan <mark.farnan at petrolink.com>wrote:

> Correct,  What I am after is a bit different.
>
> I am looking for using Websockets as native transport,  machine to machine
> running ZeroMQ natively at each end, across firewalls.   For our case we
> can't rely on a TCP port being open to work through.  Websockets provides
> the ideal transport protocol for this.
>
> i.e.
>  Connects Https
>  Handles Security (even  basic Auth is fine, as it is HttpS).
>  Upgrades the connection to Websocket  wss
>  ZeroMQ runs 'as is' over the now established Websocket.
>
> This, or something similar, is what I am thinking.   (You can actually do
> all of this in one step with a http connect and upgrade frame)
>
> So ideally a client would provide the address,  i.e.
> "wss://someserver.someplace.com/something", to the Socket Connect
> function,
> along with an optional username and PW,      It establishes the session,
> and
> ZeroMQ works as before from there.
>
> The ZeroMQ receiving the connection would need a registered callback of
> some
> kind to handle  the authentication, which just passes back 'true or false'
> if its accepted or not.
>
>
> Would anyone else maybe interested in doing this ?
>
> Regards
>
> Mark
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120706/5a03d588/attachment.htm 

From hurtonm at gmail.com  Fri Jul  6 19:40:23 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Fri, 6 Jul 2012 19:40:23 +0200
Subject: [zeromq-dev] Full-duplex tcp sockets on io threads
In-Reply-To: <CADL5_shULrk457byjqA8gSkDDCmZwDO1bv8rAir=6PZvpd0_AQ@mail.gmail.com>
References: <3bc319aab14294a9be67ee62dfc789f3@ripi.net>
	<CADL5_shULrk457byjqA8gSkDDCmZwDO1bv8rAir=6PZvpd0_AQ@mail.gmail.com>
Message-ID: <CALdZE=voMYvZ27Fq+_6we+B8MqPpbK4OBSi_0vJQtTQxbc9AAA@mail.gmail.com>

Now, both read and write operations on a given TCP socket are
performed by the same IO thread. See stream_engine implementation.

- Martin

On Fri, Jul 6, 2012 at 12:57 PM, Pieter Hintjens <ph at imatix.com> wrote:
> On Fri, Jul 6, 2012 at 5:39 PM, Rene-Pierre Lehmann <rplehmann at gmail.com> wrote:
>
>> Is it possible with zeromq to use underlying tcp sockets in a
>> full-duplex fashion?
>> By full duplex I mean that an io thread may read from a tcp socket
>> while another io thread may concurrently write to the same socket.
>
> No, this will cause the library to crash. However you can create two
> sockets in two threads, and use one for input, and one for output.
>
> -Pieter
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From mark.farnan at petrolink.com  Sat Jul  7 02:16:00 2012
From: mark.farnan at petrolink.com (Mark Farnan)
Date: Fri, 6 Jul 2012 19:16:00 -0500
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <CAJT18ibGppGntbXcvtE=vpPd52nHmfaPiikhAZ+Ky7hRq+ZiYA@mail.gmail.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>	<CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>	<043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
	<CAJT18ibGppGntbXcvtE=vpPd52nHmfaPiikhAZ+Ky7hRq+ZiYA@mail.gmail.com>
Message-ID: <017801cd5bd5$b0804390$1180cab0$@petrolink.com>

Hi Ben,

 

HttpS (more specifically SSL) is mandatory requirement in the Industry, just
forget trying to do anything without it that involves the Internet. 

 

You are correct that Websockets uses TCP underneath, just the same as HTTP
uses TCP underneath, and that's also the issue.  The protocol for Websocket
establishment works over 80/443 in places where trying to do a TCP open will
FAIL.  (we've tested it).  There are any number of statefull firewalls which
will allow the initial HTTPS connection, but will block a TCP one.

 

It will also fail a security Audit, as it is against Policy of the major
companies in question.   I'm referring here to some of the biggest Oil
companies in the world. (Policies we can't change).

 

Websockets uses standard HTTP/S methods to upgrade the connection.  We have
found issues using straight ws:// connection due to some firewalls, but so
far almost everything allows  wss://  which is SSL Websocket, upgraded from
a HTTPS connection.   It is this upgrade mechanism that is critical for it
to work.

 

Websockets are bidirectional once established, not 'client server'.    We
also don't want to use STUN, ICE or any of the Nat traversal mechanisms via
the 'cloud'.   That will also fail security audits for a start.  Initial
Establishment of the Websocket connection is client to server of course.

 

I don't personally see how adding encryption makes using something like
ZeroMQ a problem ?  If it's the fastest protocol without transport
encryption, it should remain the fastest one WITH transport layer
encryption.   

 

WS-Eventing still requires bidirectional connection establishment.  This
isn't going to fly with clients where one of the nodes is behind a NAT
device.  We looked at that, but appeared to be a non-starter.

 

 

In my opinion, Websockets will likely represent the future of Internet Push
or Publish/Subscribe Traffic (It's a hunch).  The first few messaging
systems which embrace it are likely to dominate.  I was hoping ZeroMQ might
have been going that way also.

 

RabbitMQ has a websocket transport, going to try that out next week.

 

Regards

 

Mark Farnan

Director, Product Development

Petrolink Services Inc

+1 281 908 9814

 

From: zeromq-dev-bounces at lists.zeromq.org
[mailto:zeromq-dev-bounces at lists.zeromq.org] On Behalf Of Bennie Kloosteman
Sent: Friday, July 06, 2012 10:41 AM
To: ZeroMQ development list
Subject: Re: [zeromq-dev] Websockets as a Transport ?

 

Hi Mark, 

 

Web socket uses tcp underneath so getting through the firewall is no issue
just use port 80  , and in fact web sockets ( and things that use port 80 )
have issues with some ISP html proxy servers and load balancers and it will
take time for these products to mature to handle web sockets.

 

AFAIK web sockets is also client to server and most clients already have
open all ports out and few in , if you want in a solution like Microsoft
Azure uses is ideal ( they have a tcp connection , it tries to open a
firewall port to receive and if it fails it relays it via the cloud - but in
the best case you have direct client to client connections) 

 

 I would not trust https too much it does not add a lot , make sure the
authentication is good.

 

Also  by the time you add the higher level costs like encryption what is the
point to using a light weight protocol  like zeromq ? You may as well use a
nice API like WS-Eventing  which can still do 3ms  packets  (
100Kpackets/sec per server) 

 

Ben

On Fri, Jul 6, 2012 at 6:53 AM, Mark Farnan <mark.farnan at petrolink.com>
wrote:

Correct,  What I am after is a bit different.

I am looking for using Websockets as native transport,  machine to machine
running ZeroMQ natively at each end, across firewalls.   For our case we
can't rely on a TCP port being open to work through.  Websockets provides
the ideal transport protocol for this.

i.e.
 Connects Https
 Handles Security (even  basic Auth is fine, as it is HttpS).
 Upgrades the connection to Websocket  wss
 ZeroMQ runs 'as is' over the now established Websocket.

This, or something similar, is what I am thinking.   (You can actually do
all of this in one step with a http connect and upgrade frame)

So ideally a client would provide the address,  i.e.
"wss://someserver.someplace.com/something", to the Socket Connect function,
along with an optional username and PW,      It establishes the session, and
ZeroMQ works as before from there.

The ZeroMQ receiving the connection would need a registered callback of some
kind to handle  the authentication, which just passes back 'true or false'
if its accepted or not.


Would anyone else maybe interested in doing this ?

Regards

Mark

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120706/9100c1f0/attachment.htm 

From ian.barber at gmail.com  Sat Jul  7 07:28:14 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Sat, 7 Jul 2012 06:28:14 +0100
Subject: [zeromq-dev] Using zeroMQ socket from freeradius module
In-Reply-To: <CAGb0cF=DHR3muZ1aDUhmc_FA6bOd3r82Q7r3tezOK7JZd9v7Yw@mail.gmail.com>
References: <CAGb0cFmw8KhrsLXGvO+4KZdcEeP9ffhSJD_QvZdmj18EeJXH3w@mail.gmail.com>
	<CAGb0cF=DHR3muZ1aDUhmc_FA6bOd3r82Q7r3tezOK7JZd9v7Yw@mail.gmail.com>
Message-ID: <CAEYnEwHfXn2uFZnEvu6NRJ5jsyC5ZE68knJ+RPMXkPr91Sx42A@mail.gmail.com>

On Fri, Jul 6, 2012 at 3:11 PM, Max Kuznecov <mek at mek.uz.ua> wrote:

> Oh, sorry I forgot to mention  versions:
> linux centos 6.2, zeromq 2.2.0
>
>
Is the context created in the process before the fork happens? Might be
worth trying moving where the context and socket are created - it can be
tricky with fork after context creation as the IO thread is created, the
context handle is copied with the fork, but the thread isn't.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/d9714de9/attachment.htm 

From mek at mek.uz.ua  Sat Jul  7 13:18:32 2012
From: mek at mek.uz.ua (Max Kuznecov)
Date: Sat, 7 Jul 2012 14:18:32 +0300
Subject: [zeromq-dev] Using zeroMQ socket from freeradius module
In-Reply-To: <CAEYnEwHfXn2uFZnEvu6NRJ5jsyC5ZE68knJ+RPMXkPr91Sx42A@mail.gmail.com>
References: <CAGb0cFmw8KhrsLXGvO+4KZdcEeP9ffhSJD_QvZdmj18EeJXH3w@mail.gmail.com>
	<CAGb0cF=DHR3muZ1aDUhmc_FA6bOd3r82Q7r3tezOK7JZd9v7Yw@mail.gmail.com>
	<CAEYnEwHfXn2uFZnEvu6NRJ5jsyC5ZE68knJ+RPMXkPr91Sx42A@mail.gmail.com>
Message-ID: <CAGb0cFmjbuRUYaO8b6gQ9udnn+tLC9WpNQbn5Ti+Mbsa+wK7Xg@mail.gmail.com>

Indeed! Looks like freeradius instantiates its modules before forking.
I've postponed context/socket initialization and now everything works perfectly.

Thank you very much!

2012/7/7 Ian Barber <ian.barber at gmail.com>:
>
>
> On Fri, Jul 6, 2012 at 3:11 PM, Max Kuznecov <mek at mek.uz.ua> wrote:
>>
>> Oh, sorry I forgot to mention  versions:
>> linux centos 6.2, zeromq 2.2.0
>>
>
> Is the context created in the process before the fork happens? Might be
> worth trying moving where the context and socket are created - it can be
> tricky with fork after context creation as the IO thread is created, the
> context handle is copied with the fork, but the thread isn't.
>
> Ian
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
~syhpoon

From mark_sutheran at yahoo.com  Sat Jul  7 15:42:42 2012
From: mark_sutheran at yahoo.com (Mark Sutheran)
Date: Sat, 7 Jul 2012 06:42:42 -0700 (PDT)
Subject: [zeromq-dev] Asymmetrical PUB-SUB problem
Message-ID: <1341668562.77211.YahooMailNeo@web160306.mail.bf1.yahoo.com>

I'm having a strange?asymmetrical?problem with tcp PUB/SUB between two boxes - it appears to fail one way.



Details:?
* zeromq version: 3.2, using the Java wrapper.
* Scenario has two machines A (Ubuntu, x64) and B (Angstrom, arm8)
* Both connected directly via wireless interface (B is acting as AP).

Scenario 1: A->B. All as expected.
* Box A:PUB bind wlan0 - publish messages 1x per second.
*?Box?A:SUB connect <ipx> - see messages, yay!
*?Box?B:SUB connect <ipx> - see messages, yay!

Scenario 2: B->A. Problem
*?Box?B:PUB bind wlan0 - publish messages 1x per second.
*?Box?B:SUB connect <ipy> - see messages, yay!
*?Box?A:SUB connect <ipx> - no messages, boo!

However, if I change pattern...

Scenario 3: Comparison
*?Box?A: REQ - make n requests.
*?Box?B: REP - send n replies. Replies received on Machine A as expected. Yay.

There are absolutely no messages received at the SUB on A from the PUB on B (irrespective of the process start order) however, connecting a SUB on B to the interface's ip works fine. If I reverse the boxes I see the behaviour I expect - i.e. the SUBs on both boxes see the messages. As a contrast: if I run a REQ on A and REP on B, everything works as expected and the message sent from B is received at A.

This is presumably a problem with my hardware/OS but I'm not sure even where to start looking... any pointers would be greatly appreciated.

Cheers,
Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/aab2fab0/attachment.htm 

From jhawk28 at gmail.com  Sat Jul  7 16:14:32 2012
From: jhawk28 at gmail.com (Joshua Foster)
Date: Sat, 07 Jul 2012 10:14:32 -0400
Subject: [zeromq-dev] Asymmetrical PUB-SUB problem
In-Reply-To: <1341668562.77211.YahooMailNeo@web160306.mail.bf1.yahoo.com>
References: <1341668562.77211.YahooMailNeo@web160306.mail.bf1.yahoo.com>
Message-ID: <4FF84448.6080306@gmail.com>

Try binding to * on box b when publishing.

Joshua

> Mark Sutheran <mailto:mark_sutheran at yahoo.com>
> July 7, 2012 9:42 AM
> I'm having a strange asymmetrical problem with tcp PUB/SUB between two 
> boxes - it appears to fail one way.
>
> Details:
> * zeromq version: 3.2, using the Java wrapper.
> * Scenario has two machines A (Ubuntu, x64) and B (Angstrom, arm8)
> * Both connected directly via wireless interface (B is acting as AP).
>
> Scenario 1: A->B. All as expected.
> * Box A:PUB bind wlan0 - publish messages 1x per second.
> * Box A:SUB connect <ipx> - see messages, yay!
> * Box B:SUB connect <ipx> - see messages, yay!
>
> Scenario 2: B->A. Problem
> * Box B:PUB bind wlan0 - publish messages 1x per second.
> * Box B:SUB connect <ipy> - see messages, yay!
> * Box A:SUB connect <ipx> - no messages, boo!
>
> However, if I change pattern...
>
> Scenario 3: Comparison
> * Box A: REQ - make n requests.
> * Box B: REP - send n replies. Replies received on Machine A as 
> expected. Yay.
>
> There are absolutely no messages received at the SUB on A from the PUB 
> on B (irrespective of the process start order) however, connecting a 
> SUB on B to the interface's ip works fine. If I reverse the boxes I 
> see the behaviour I expect - i.e. the SUBs on both boxes see the 
> messages. As a contrast: if I run a REQ on A and REP on B, everything 
> works as expected and the message sent from B is received at A.
>
> This is presumably a problem with my hardware/OS but I'm not sure even 
> where to start looking... any pointers would be greatly appreciated.
>
> Cheers,
> Mark
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/9890e689/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: compose-unknown-contact.jpg
Type: image/jpeg
Size: 770 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/9890e689/attachment.jpg 

From mark_sutheran at yahoo.com  Sat Jul  7 16:36:26 2012
From: mark_sutheran at yahoo.com (Mark Sutheran)
Date: Sat, 7 Jul 2012 07:36:26 -0700 (PDT)
Subject: [zeromq-dev] Asymmetrical PUB-SUB problem
In-Reply-To: <4FF84448.6080306@gmail.com>
References: <1341668562.77211.YahooMailNeo@web160306.mail.bf1.yahoo.com>
	<4FF84448.6080306@gmail.com>
Message-ID: <1341671786.85047.YahooMailNeo@web160305.mail.bf1.yahoo.com>

Thanks for the suggestion Joshua. Unfortunately this doesn't fix the problem.

(I restricted the publisher to the one interface in order to avoid any possible confusion due to?potential?routing through some other interface)

However to add to the information - the SUB on box B can successfully connect to and receive data from any valid interface (wlan0, localhost, eth0) when the PUB binds to *.

Cheers,

Mark


________________________________
 From: Joshua Foster <jhawk28 at gmail.com>
To: Mark Sutheran <mark_sutheran at yahoo.com>; ZeroMQ development list <zeromq-dev at lists.zeromq.org> 
Sent: Saturday, 7 July 2012, 22:14
Subject: Re: [zeromq-dev] Asymmetrical PUB-SUB problem
 

Try binding to * on box b 
when publishing.

Joshua


Mark Sutheran
>July 7, 2012 9:42 AM
>I'm having a strange?asymmetrical?problem with tcp 
PUB/SUB between two boxes - it appears to fail one way.
>
>
>
>Details:?
>* zeromq version: 3.2, 
using the Java wrapper.
>* Scenario has two machines A (Ubuntu, x64) and B 
(Angstrom, arm8)
>* Both connected directly via wireless interface (B is acting as AP).
>
>
>Scenario 1: A->B. All as expected.
>* Box A:PUB bind wlan0 - 
publish messages 1x per second.
>*?Box?A:SUB connect <ipx> - see 
messages, yay!
>*?Box?B:SUB connect <ipx> - see messages, 
yay!
>
>
>Scenario 2: B->A. 
Problem
>*?Box?B:PUB bind wlan0 - publish messages 1x per second.
>*?Box?B:SUB connect 
<ipy> - see messages, yay!
>*?Box?A:SUB connect <ipx> - no 
messages, boo!
>
>
>However, if I change pattern...
>
>
>Scenario 3: Comparison
>*?Box?A: REQ - make n 
requests.
>*?Box?B: REP - send n replies. Replies received on Machine A as expected. Yay.
>
>
>There are absolutely no messages received at the 
SUB on A from the PUB on B (irrespective of the process start order) 
however, connecting a SUB on B to the interface's ip works fine. If I reverse the boxes I see the 
behaviour I expect - i.e. the SUBs on both boxes see the messages. As a 
contrast: if I run a REQ on A and REP on B, everything works as expected and the message sent from B is received at A.
>
>
>This is presumably a 
problem with my hardware/OS but I'm not sure even where to start 
looking... any pointers would be greatly appreciated.
>
>
>Cheers,
>Mark
>
>
>_______________________________________________
>zeromq-dev
 mailing list
>zeromq-dev at lists.zeromq.org
>http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/3deea357/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: compose-unknown-contact.jpg
Type: image/jpeg
Size: 770 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/3deea357/attachment.jpg 

From bklooste at gmail.com  Sat Jul  7 17:48:05 2012
From: bklooste at gmail.com (Bennie Kloosteman)
Date: Sat, 7 Jul 2012 23:48:05 +0800
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <017801cd5bd5$b0804390$1180cab0$@petrolink.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
	<CAEYnEwFOp07bwLmrJcotCY-xZtu2tDyk79K-10hvYf6esW01gw@mail.gmail.com>
	<CAA0gF6ob+JpT-4Thtgo3COHDtO22q-JLmnchpg1whiJJoQcOnQ@mail.gmail.com>
	<043801cd5b00$f84d8550$e8e88ff0$@petrolink.com>
	<CAJT18ibGppGntbXcvtE=vpPd52nHmfaPiikhAZ+Ky7hRq+ZiYA@mail.gmail.com>
	<017801cd5bd5$b0804390$1180cab0$@petrolink.com>
Message-ID: <CAJT18ib0p1ufo+2ZboCj5jExoggNZwNUD=iWE5BWCUXDYfedew@mail.gmail.com>

On Sat, Jul 7, 2012 at 8:16 AM, Mark Farnan <mark.farnan at petrolink.com>wrote:

> Hi Ben,****
>
> ** **
>
> HttpS (more specifically SSL) is mandatory requirement in the Industry,
> just forget trying to do anything without it that involves the Internet.
>

I know but  you said minimal authentication  , if you break the
authentication you can create an SSL socket.
I have also wondered about HTTPS  on private network  , people with network
access have admin and can watch your packets  and on the internet it is
really hard to separate a single http channel in a gigabit traffic stream (
I tried )... It does cover low speed ISPs and wireless connections where
people don't turn on encryption and that makes it essential .

> ****
>
> ** **
>
> You are correct that Websockets uses TCP underneath, just the same as HTTP
> uses TCP underneath, and that?s also the issue.  The protocol for Websocket
> establishment works over 80/443 in places where trying to do a TCP open
> will FAIL.  (we?ve tested it).  There are any number of statefull firewalls
> which will allow the initial HTTPS connection, but will block a TCP one.
>

So are they look for a http in the header ? That's easy enough to dummy ...
I have frequently used tcp over port 80 which works through most firewalls
 unless you have a nasty http proxy server / load balancer but a lot of
those ( not all ) also have issues with web sockets.


> ****
>
> ** **
>
> It will also fail a security Audit, as it is against Policy of the major
> companies in question.   I?m referring here to some of the biggest Oil
> companies in the world. (Policies we can?t change).
>

How can it fail when web sockets do the same thing ..I worked for a major
merchant bank and if you had a good reason security was normally granted.
You may be stuck with a requirement for encryption which can make sense,


> ****
>
> ** **
>
> Websockets uses standard HTTP/S methods to upgrade the connection.  We
> have found issues using straight ws:// connection due to some firewalls,
> but so far almost everything allows  wss://  which is SSL Websocket,
> upgraded from a HTTPS connection.   It is this upgrade mechanism that is
> critical for it to work.
>

That is useful a friend had a look at it last year and had some hassles.
That said unless you have tested MANY ISP I would  extremely careful with a
new technology.   especially transparent  http proxies , load balancers
which can be 10 year old boxes and nasty things like the great firewall. I
mean some isp love breaking long connections since you may be hogging
resources.

> ****
>
> ** **
>
> Websockets are bidirectional once established, not ?client server?.    We
> also don?t want to use STUN, ICE or any of the Nat traversal mechanisms via
> the ?cloud?.   That will also fail security audits for a start.  Initial
> Establishment of the Websocket connection is client to server of course.
>

Its still client server even if bi directional eg you cant have 2 clients
communicating directly without going to your server.   Its just the client
is piggy backing on the return channel which makes it similar to bi
directional but unlike real bi directional system if the one socket goes
the connection goes.

> ** **
>
> I don?t personally see how adding encryption makes using something like
> ZeroMQ a problem ?  If it?s the fastest protocol without transport
> encryption, it should remain the fastest one WITH transport layer
> encryption.
>

Once you add the encryption and internet latency the difference becomes
very small possibly negligible.
****

> ** **
>
> WS-Eventing still requires bidirectional connection establishment.  This
> isn?t going to fly with clients where one of the nodes is behind a NAT
> device.  We looked at that, but appeared to be a non-starter.****
>
> **
>
Nope ,nothing in the spec says that only the common implementations are bi
directional .I have seen store and forward  and I  built a system with
WS-Eventing where the clients  ( 3000 windows mobile handhelds behind many
different  NATs) polled  a message holding service which held the eventing
server for new messages.  Building push was also possible  , since it was
WCF based I quickly build a prototype ( change the binding and hook it up
to where the polled messages were dispatched ) which automatically used  a
connection similar to web sockets but it didn't make it because we had many
connection drop outs(mobile nets)  and there were a few bugs to sort out
which would make it too expensive .


> **
>
> ** **
>
> In my opinion, Websockets will likely represent the future of Internet
> Push or Publish/Subscribe Traffic (It?s a hunch).  The first few messaging
> systems which embrace it are likely to dominate.  I was hoping ZeroMQ might
> have been going that way also.
>

I really have seen little use for these roles, to me the compelling roll
for web sockets is to write more complex html5 javascript apps that use
them due to a simple API will less security hassles . Not to replace
messaging systems which all have ways through firewalls already.

For many cases the cloud based connection services offer a compelling
argument due to the possibility of server less communication for many
clients and a fall back to relay.  And for internet based system a huge
issue in communication is still the protocol hence FIX and SOA /WSEventing
wsdl systems have a big place.  Many messaging systems like zeroMQ are used
in merchant banks and things like that and they are counting every tenth of
a milli second especially for   price information and frequent trading
systems.  In many of those environments they use private links so no need
for encryption. So to me its likely that messaging will remain fragmented.
I thought the same about WS-Eventing and to me its a great system , very
flexible ( eg easy to hook in many systems , or down load events to new
servers to handle alsmots limitless load ) but few people have used it -
public internet pub sub itsefl is not that popular, both of which to me are
dissapointing.

> ****
>
> ** **
>
> RabbitMQ has a websocket transport, going to try that out next week.
>

It makes more sense for AMQP based systems  , they are more complex/
flexible  and less light weight. I havent heard anyone in ZeroMq using http
yet alone https though im only  a casual observer now.



Regards,

Ben
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/0e6afa9e/attachment.htm 

From mark_sutheran at yahoo.com  Sat Jul  7 18:22:25 2012
From: mark_sutheran at yahoo.com (Mark Sutheran)
Date: Sat, 7 Jul 2012 09:22:25 -0700 (PDT)
Subject: [zeromq-dev] Asymmetrical PUB-SUB problem
In-Reply-To: <1341671786.85047.YahooMailNeo@web160305.mail.bf1.yahoo.com>
References: <1341668562.77211.YahooMailNeo@web160306.mail.bf1.yahoo.com>
	<4FF84448.6080306@gmail.com>
	<1341671786.85047.YahooMailNeo@web160305.mail.bf1.yahoo.com>
Message-ID: <1341678145.8240.YahooMailNeo@web160302.mail.bf1.yahoo.com>

Ok, apols - looks like it's my bad!

Rebuilt all the libs on the client box (A) to the latest and greatest and the problem went away. So I guess this was a backwards-only?compatibility?issue between different versions (fits the pattern)

Cheers,
Mark


________________________________
 From: Mark Sutheran <mark_sutheran at yahoo.com>
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org> 
Sent: Saturday, 7 July 2012, 22:36
Subject: Re: [zeromq-dev] Asymmetrical PUB-SUB problem
 

Thanks for the suggestion Joshua. Unfortunately this doesn't fix the problem.

(I restricted the publisher to the one interface in order to avoid any possible confusion due to?potential?routing through some other interface)

However to add to the information - the SUB on box B can successfully connect to and receive data from any valid interface (wlan0, localhost, eth0) when the PUB binds to *.

Cheers,

Mark


________________________________
 From: Joshua Foster <jhawk28 at gmail.com>
To: Mark Sutheran <mark_sutheran at yahoo.com>; ZeroMQ development list <zeromq-dev at lists.zeromq.org> 
Sent: Saturday, 7 July 2012, 22:14
Subject: Re: [zeromq-dev] Asymmetrical PUB-SUB problem
 

Try binding to * on box b 
when publishing.

Joshua


Mark Sutheran
>July 7, 2012 9:42 AM
>I'm having a strange?asymmetrical?problem with tcp 
PUB/SUB between two boxes - it appears to fail one way.
>
>
>
>Details:?
>* zeromq version: 3.2, 
using the Java wrapper.
>* Scenario has two machines A (Ubuntu, x64) and B 
(Angstrom, arm8)
>* Both connected directly via wireless interface (B is acting as AP).
>
>
>Scenario 1: A->B. All as expected.
>* Box A:PUB bind wlan0 - 
publish messages 1x per second.
>*?Box?A:SUB connect <ipx> - see 
messages, yay!
>*?Box?B:SUB connect <ipx> - see messages, 
yay!
>
>
>Scenario 2: B->A. 
Problem
>*?Box?B:PUB bind wlan0 - publish messages 1x per second.
>*?Box?B:SUB connect 
<ipy> - see messages, yay!
>*?Box?A:SUB connect <ipx> - no 
messages, boo!
>
>
>However, if I change pattern...
>
>
>Scenario 3: Comparison
>*?Box?A: REQ - make n 
requests.
>*?Box?B: REP - send n replies. Replies received on Machine A as expected. Yay.
>
>
>There are absolutely no messages received at the 
SUB on A from the PUB on B (irrespective of the process start order) 
however, connecting a SUB on B to the interface's ip works fine. If I reverse the boxes I see the 
behaviour I expect - i.e. the SUBs on both boxes see the messages. As a 
contrast: if I run a REQ on A and REP on B, everything works as expected and the message sent from B is received at A.
>
>
>This is presumably a 
problem with my hardware/OS but I'm not sure even where to start 
looking... any pointers would be greatly appreciated.
>
>
>Cheers,
>Mark
>
>
>_______________________________________________
>zeromq-dev
 mailing list
>zeromq-dev at lists.zeromq.org
>http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev



_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/8b0301ff/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: compose-unknown-contact.jpg
Type: image/jpeg
Size: 770 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120707/8b0301ff/attachment.jpg 

From Sharon.Ben-Asher at avg.com  Sun Jul  8 08:40:04 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Sun, 8 Jul 2012 06:40:04 +0000
Subject: [zeromq-dev] Error: comparison between signed and unsigned
 integer expressions
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FC024@ex10czdmb01>

Hi Pieter,

Here you go
Issue LIBZMQ-394 - socklen_t is signed int in Android NDK

Now, I will be glad for some pointer on my linkage error ... thanks :)

Sharon


From: Pieter Hintjens <ph at imatix.com>
Sent: Thu, 5 Jul 2012 19:02:33 +0900
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org>
Subject: Re: [zeromq-dev] Error: comparison between signed and unsigned integer expressions

Hi Sharon,

OK, so socklen_t is definitely signed on your platform. If you would make an issue in Jira, I'll make a patch for all current 0MQ versions.

Thanks
Pieter
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120708/ad2b26b0/attachment.htm 

From Sharon.Ben-Asher at avg.com  Sun Jul  8 10:35:39 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Sun, 8 Jul 2012 08:35:39 +0000
Subject: [zeromq-dev] Error: cannot link with -lzmq
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FD08E@ex10czdmb01>

Ok,

I solved the issue by installing gcc-multilib and g++-multilib packages.
Maybe you should mention this in the Android build webpage?

Thanks,

Sharon

From: Sharon Ben-Asher
Sent: Thursday, July 05, 2012 1:29 PM
To: 'zeromq-dev at lists.zeromq.org'
Subject: Error: cannot link with -lzmq

Hello again,

After successful build of the C lib, I have a problem with building zmq.jar:

configure:15622: checking for zmq_init in -lzmq
configure:15647: g++ -o conftest -g -O2 -D_REENTRANT -D_THREAD_SAFE -fPIC -I/home/sharon/zeromq-android/include  -L/home/sharon/zeromq-android/lib conftest.cpp -lzmq   >&5
/usr/bin/ld: skipping incompatible /home/sharon/zeromq-android/lib/libzmq.so when searching for -lzmq
/usr/bin/ld: skipping incompatible /home/sharon/zeromq-android/lib/libzmq.a when searching for -lzmq
/usr/bin/ld: cannot find -lzmq
collect2: ld returned 1 exit status

Thanks

Sharon


From: Sharon Ben-Asher
Sent: Thursday, July 05, 2012 12:56 PM
To: 'zeromq-dev at lists.zeromq.org'
Subject: RE: Error: comparison between signed and unsigned integer expressions

I solved the problem by converting the sizeof to int like this
if (sa->sa_family == AF_INET && sa_len >= (int)sizeof (address.ipv4)) {
I guess it's a platform specific issue? Anyway, it is needed to compile the lib for ARM.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120708/c5dcca1d/attachment.htm 

From ph at imatix.com  Mon Jul  9 02:42:51 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Mon, 9 Jul 2012 09:42:51 +0900
Subject: [zeromq-dev] Error: cannot link with -lzmq
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B2E7FD08E@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B2E7FD08E@ex10czdmb01>
Message-ID: <CADL5_si3TMSdKa5zHBN=Mjguqr+47eambBbiJEEA_escdOz31Q@mail.gmail.com>

Hi Sharon,

> I solved the issue by installing gcc-multilib and g++-multilib packages.
> Maybe you should mention this in the Android build webpage?

It's a wiki, go ahead and edit that page.

-Pieter

From ph at imatix.com  Mon Jul  9 03:41:26 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Mon, 9 Jul 2012 10:41:26 +0900
Subject: [zeromq-dev] Weird Fish Book
Message-ID: <CADL5_shxzbsoPaG3URwphARRuCN5TGm9-YY_3nVUOOUVaXHcRA@mail.gmail.com>

Hi all,

FYI, the 0MQ Weird Fish Book (based on the Guide) is coming out in
paper sometime soon, and is available in Safari:
http://my.safaribooksonline.com/book/-/9781449334437

Thanks to the many people who contributed examples, errata, and suggestions!

-Pieter

From thatch45 at gmail.com  Mon Jul  9 06:53:24 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Sun, 8 Jul 2012 22:53:24 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
Message-ID: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>

Hi,

I have a report of an assertion in Salt. It is very rare and I cannot
reliably reproduce it. The user that has reported it gave me a strace
output which makes it look like the assertion is happening in a
multiprocess. I am aware that you cannot have the same socket call from
another thread or process, but I see no indication that this is the case.

Here is the trace:

14:51:29.164045 epoll_wait(9, {{EPOLLIN, {u32=44624800, u64=44624800}}},
256, 4294967295) = 1
14:51:55.130374 poll([{fd=8, events=POLLIN}], 1, 0) = 1 ([{fd=8,
revents=POLLIN}])
14:51:55.130428 write(2, "Assertion failed: ok (mailbox.cpp:84)\n", 38) =
-1 EBADF (Bad file descriptor)
14:51:55.130466 rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0
14:51:55.130500 tgkill(5969, 5980, SIGABRT) = 0
14:51:55.130530 --- SIGABRT (Aborted) @ 0 (0) ---

Any ideas?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120708/86789c79/attachment.htm 

From Lev at visionmap.com  Mon Jul  9 09:49:12 2012
From: Lev at visionmap.com (Lev Borovoi)
Date: Mon, 9 Jul 2012 10:49:12 +0300
Subject: [zeromq-dev] "Hello world" example crashes on 3.2.0
Message-ID: <E15948BC84BC624E822DC6493E0F48BE32B2DB9172@MARS.visionmap.co.il>

Hi

I tried to run the "Hello world" server source (see http://zguide.zeromq.org/page:all#toc5). Since I want to use 0mq 3.2.0 rather than 2.2.0, I replaced zmq_recv by zmq_recvmsg and zmq_send by zmq_sendmsg. However, when I start the client, the server crashes with this message:

Assertion failed: (msg_->flags () & msg_t::identity) == 0 (..\..\..\src\router.cpp:222)

Since I'm running this on Windows, I also replaced sleep(1) with Sleep(1000).

What am I doing wrong and how should I fix it?

This email and any files transmitted with it are confidential and contain proprietary information belonging to VisionMap Ltd. VisionMap Ltd. asserts in respect of this email and any files transmitted with it all rights for confidentiality and proprietary interests to the fullest extent permitted by law. If you are not the intended recipient you are notified that disclosing, copying, distributing or taking any action in reliance on the contents of this information is strictly prohibited. Thank you. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/b64254cb/attachment.htm 

From ph at imatix.com  Mon Jul  9 10:40:45 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Mon, 9 Jul 2012 17:40:45 +0900
Subject: [zeromq-dev] "Hello world" example crashes on 3.2.0
In-Reply-To: <E15948BC84BC624E822DC6493E0F48BE32B2DB9172@MARS.visionmap.co.il>
References: <E15948BC84BC624E822DC6493E0F48BE32B2DB9172@MARS.visionmap.co.il>
Message-ID: <CADL5_sgkDTXN+Dpu_gPrp1QL2pdM5w-5n7jh5KU+DSdQuUq2yg@mail.gmail.com>

Lev,

> I tried to run the ?Hello world? server source (see
> http://zguide.zeromq.org/page:all#toc5). Since I want to use 0mq 3.2.0
> rather than 2.2.0, I replaced zmq_recv by zmq_recvmsg and zmq_send by
> zmq_sendmsg. However, when I start the client, the server crashes with this
> message:
>
> Assertion failed: (msg_->flags () & msg_t::identity) == 0
> (..\..\..\src\router.cpp:222)

I could not reproduce this. Could you send the actual client and
server code that you used?

Thanks
Pieter

From ph at imatix.com  Mon Jul  9 10:52:45 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Mon, 9 Jul 2012 17:52:45 +0900
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
Message-ID: <CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>

Hi Thomas,

Can you definitely exclude that sockets end up in the wrong thread? I
mean, are you at any time passing socket handles between threads?

-Pieter

On Mon, Jul 9, 2012 at 1:53 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:
> Hi,
>
> I have a report of an assertion in Salt. It is very rare and I cannot
> reliably reproduce it. The user that has reported it gave me a strace output
> which makes it look like the assertion is happening in a multiprocess. I am
> aware that you cannot have the same socket call from another thread or
> process, but I see no indication that this is the case.
>
> Here is the trace:
>
> 14:51:29.164045 epoll_wait(9, {{EPOLLIN, {u32=44624800, u64=44624800}}},
> 256, 4294967295) = 1
> 14:51:55.130374 poll([{fd=8, events=POLLIN}], 1, 0) = 1 ([{fd=8,
> revents=POLLIN}])
> 14:51:55.130428 write(2, "Assertion failed: ok (mailbox.cpp:84)\n", 38) = -1
> EBADF (Bad file descriptor)
> 14:51:55.130466 rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0
> 14:51:55.130500 tgkill(5969, 5980, SIGABRT) = 0
> 14:51:55.130530 --- SIGABRT (Aborted) @ 0 (0) ---
>
> Any ideas?
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From rblists at gmail.com  Mon Jul  9 11:40:01 2012
From: rblists at gmail.com (Raphael Bauduin)
Date: Mon, 9 Jul 2012 11:40:01 +0200
Subject: [zeromq-dev] Interrupted System Call: advice to handle it
In-Reply-To: <01B73EA7-9C85-467A-AF3E-533BC52E5945@chuckremes.com>
References: <CAONrwUHbDUGwiOz-_Qc81m+Vn+uninnUjioSar8_NDO4JZTH3w@mail.gmail.com>
	<01B73EA7-9C85-467A-AF3E-533BC52E5945@chuckremes.com>
Message-ID: <CAONrwUE7CvQORxF7yXnGb2Cwf2S8ak+tDPHYaBxfBh9H=TXhEw@mail.gmail.com>

On Wed, Jul 4, 2012 at 5:20 PM, Chuck Remes <lists at chuckremes.com> wrote:
>
> On Jul 4, 2012, at 5:05 AM, Raphael Bauduin wrote:
>
>> Hi,
>>
>> I'm using the ruby zmq bindings in a web application. I regularly get
>> error message "ZMQ::Error: Interrupted system call" related to a send.
>> This is in a Ruby on Rails application served with passenger, which
>> spawns worker processes. I think I have identified a process that
>> generated this error, and an strace on it shows no activity at all.
>> This process however keeps open a connection to the mysql server. An
>> accumulation of such errors will eventually become problematic server
>> side, in addition to clients getting an error page and messages being
>> lost.
>
> I'm assuming this happens under MRI. Is it 1.8.x or 1.9.x?

It is REE: ruby 1.8.7 (2012-02-08 MBARI 8/0x6770 on patchlevel 358)
[x86_64-linux], MBARI 0x6770, Ruby Enterprise Edition 2012.02

>
> Do you see the same behavior when running your app with JRuby or Rubinius?
>

The problem is that I don't have the problem systematically. It
happens once every x days in production where there are thousands of
page views that run the code in question. So it's very hard to
reproduce.

>> I'm looking for advice in avoiding this error and possibly for further
>> debugging hints. Related to that I have several questions:
>> - Should I simply catch this exception, and retry the send if needed?
>> As this is done in the process sending the page content back to the
>> client, won't it possibly make some requests too slow? (This could
>> still be better than an error as we have currently)
>
> Using exception handling for flow control in Ruby can be slow. But unless you are building the next amazon.com then it probably won't hurt you too much. You could give this a try though it's always better to figure out the actual underlying cause and fix it. Using exceptions here is just a band-aid.
>
>> - If my understanding is correct, the problem occurs with blocking
>> syscalls, and requests having the error don't return any content to
>> the client. But what happens if I make the send non blocking?
>> (http://zeromq.github.com/rbzmq/classes/ZMQ/Socket.html#M000010)
>
> Try it and see.

My question was more about knowing if the same problem could occur. As
mentioned above, I can't reproduce the problem systematically.

>
>> - Finally, what might interrupt the syscall? Any interesting read about this?
>
> Something in your app is generating a signal. The technique I use to figure out these kinds of errors is to run my app under other Ruby runtimes. Most of the time they will fail differently and/or give me an exact backtrace pointing to the source of the problem.

Can it also be a signal coming from outside the app, eg passenger?

Or can it be due to the fact that I set the LINGER option?
  s.setsockopt(ZMQ::LINGER,100)
  ..
  s.send(m)
  s.close

Any suggestion on this would be really welcome!

>
> Lastly, you may want to look at the ffi-rzmq gem (disclaimer: I'm its maintainer). It has a different API from the zmq gem but it appears to enjoy wider usage by the community so it may be a bit more stable.

Thanks for the tip, I add it as an option, but I'd like to understand
what's going on too.

Raph


>
> cr
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev



-- 
Web database: http://www.myowndb.com
Free Software Developers Meeting: http://www.fosdem.org

From thatch45 at gmail.com  Mon Jul  9 20:13:09 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Mon, 9 Jul 2012 12:13:09 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
Message-ID: <CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>

No, I am confident that we are not passing sockets through threads or
processes.
Could this have anything to do with closing and re opening connections?

On Mon, Jul 9, 2012 at 2:52 AM, Pieter Hintjens <ph at imatix.com> wrote:

> Hi Thomas,
>
> Can you definitely exclude that sockets end up in the wrong thread? I
> mean, are you at any time passing socket handles between threads?
>
> -Pieter
>
> On Mon, Jul 9, 2012 at 1:53 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:
> > Hi,
> >
> > I have a report of an assertion in Salt. It is very rare and I cannot
> > reliably reproduce it. The user that has reported it gave me a strace
> output
> > which makes it look like the assertion is happening in a multiprocess. I
> am
> > aware that you cannot have the same socket call from another thread or
> > process, but I see no indication that this is the case.
> >
> > Here is the trace:
> >
> > 14:51:29.164045 epoll_wait(9, {{EPOLLIN, {u32=44624800, u64=44624800}}},
> > 256, 4294967295) = 1
> > 14:51:55.130374 poll([{fd=8, events=POLLIN}], 1, 0) = 1 ([{fd=8,
> > revents=POLLIN}])
> > 14:51:55.130428 write(2, "Assertion failed: ok (mailbox.cpp:84)\n", 38)
> = -1
> > EBADF (Bad file descriptor)
> > 14:51:55.130466 rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0
> > 14:51:55.130500 tgkill(5969, 5980, SIGABRT) = 0
> > 14:51:55.130530 --- SIGABRT (Aborted) @ 0 (0) ---
> >
> > Any ideas?
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/d05c1cd9/attachment.htm 

From ian.barber at gmail.com  Mon Jul  9 20:52:05 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Mon, 9 Jul 2012 19:52:05 +0100
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
Message-ID: <CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>

On Mon, Jul 9, 2012 at 7:13 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:

> No, I am confident that we are not passing sockets through threads or
> processes.
> Could this have anything to do with closing and re opening connections?
>

Do you pull the FD out directly, or is this just from accessing via the
regular ZMQ_ functions? Also, do you have any idea how quickly a socket
might be destroyed and recreated or similar?

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/20fd83ff/attachment.htm 

From thatch45 at gmail.com  Mon Jul  9 20:58:11 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Mon, 9 Jul 2012 12:58:11 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
Message-ID: <CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>

On Mon, Jul 9, 2012 at 12:52 PM, Ian Barber <ian.barber at gmail.com> wrote:

>
>
> On Mon, Jul 9, 2012 at 7:13 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:
>
>> No, I am confident that we are not passing sockets through threads or
>> processes.
>> Could this have anything to do with closing and re opening connections?
>>
>
> Do you pull the FD out directly, or is this just from accessing via the
> regular ZMQ_ functions? Also, do you have any idea how quickly a socket
> might be destroyed and recreated or similar?
>
> Ian
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
There is an option in Salt that makes the SUB socket close and reconnect to
the master every N seconds to recover from disconnection over unreliable
connections, aka the internet. The default is every 60 seconds

It just uses the regular ZMQ function via pyzmq
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/72dea619/attachment.htm 

From lists at chuckremes.com  Mon Jul  9 21:14:24 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 9 Jul 2012 14:14:24 -0500
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
Message-ID: <D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>


On Jul 9, 2012, at 1:58 PM, Thomas S Hatch wrote:

> On Mon, Jul 9, 2012 at 12:52 PM, Ian Barber <ian.barber at gmail.com> wrote:
> 
> 
> On Mon, Jul 9, 2012 at 7:13 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:
> No, I am confident that we are not passing sockets through threads or processes.
> Could this have anything to do with closing and re opening connections?
> 
> Do you pull the FD out directly, or is this just from accessing via the regular ZMQ_ functions? Also, do you have any idea how quickly a socket might be destroyed and recreated or similar? 
> 
> Ian
> 
> 
> There is an option in Salt that makes the SUB socket close and reconnect to the master every N seconds to recover from disconnection over unreliable connections, aka the internet. The default is every 60 seconds
> 
> It just uses the regular ZMQ function via pyzmq

Can we assume that this socket is removed from the poll items list that zmq_poll() uses? (I am assuming you are using zmq_poll() in Salt and that these SUB sockets are added/deleted as necessary.)

cr


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/23e20240/attachment.htm 

From thatch45 at gmail.com  Mon Jul  9 21:17:06 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Mon, 9 Jul 2012 13:17:06 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
Message-ID: <CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>

On Mon, Jul 9, 2012 at 1:14 PM, Chuck Remes <lists at chuckremes.com> wrote:

>
> On Jul 9, 2012, at 1:58 PM, Thomas S Hatch wrote:
>
> On Mon, Jul 9, 2012 at 12:52 PM, Ian Barber <ian.barber at gmail.com> wrote:
>
>>
>>
>> On Mon, Jul 9, 2012 at 7:13 PM, Thomas S Hatch <thatch45 at gmail.com>wrote:
>>
>>> No, I am confident that we are not passing sockets through threads or
>>> processes.
>>> Could this have anything to do with closing and re opening connections?
>>>
>>
>> Do you pull the FD out directly, or is this just from accessing via the
>> regular ZMQ_ functions? Also, do you have any idea how quickly a socket
>> might be destroyed and recreated or similar?
>>
>> Ian
>>
>>
> There is an option in Salt that makes the SUB socket close and reconnect
> to the master every N seconds to recover from disconnection over unreliable
> connections, aka the internet. The default is every 60 seconds
>
> It just uses the regular ZMQ function via pyzmq
>
>
> Can we assume that this socket is removed from the poll items list that
> zmq_poll() uses? (I am assuming you are using zmq_poll() in Salt and that
> these SUB sockets are added/deleted as necessary.)
>
> cr
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
> Yes, they are, here is the code:
https://github.com/saltstack/salt/blob/develop/salt/minion.py#L470

I do not doubt that I have made a mistake in here
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/fead149c/attachment.htm 

From lists at chuckremes.com  Mon Jul  9 21:43:08 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 9 Jul 2012 14:43:08 -0500
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
	<CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
Message-ID: <B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>

On Jul 9, 2012, at 2:17 PM, Thomas S Hatch wrote:

> Yes, they are, here is the code:
> https://github.com/saltstack/salt/blob/develop/salt/minion.py#L470
> 
> I do not doubt that I have made a mistake in here

I see that you are unregistering, closing the socket, and then immediately creating a new socket with the same IDENTITY. 

I did not see any code setting the LINGER option to drop packets, so if a late packet arrives I suppose it could cause that zmq_close() to hang. 

Alternately, reusing the IDENTITY before the socket is completely closed (an async operation) may be a cause. Any chance you can close the socket and then schedule (within your loop) the creation of a replacement socket a few milliseconds later?

These are just two guesses based on 5m review of the code. I hope it sparks an idea or a new avenue to pursue.

cr
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/4e400880/attachment.htm 

From thatch45 at gmail.com  Mon Jul  9 21:59:54 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Mon, 9 Jul 2012 13:59:54 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
	<CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
	<B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>
Message-ID: <CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>

On Mon, Jul 9, 2012 at 1:43 PM, Chuck Remes <lists at chuckremes.com> wrote:

> On Jul 9, 2012, at 2:17 PM, Thomas S Hatch wrote:
>
> Yes, they are, here is the code:
> https://github.com/saltstack/salt/blob/develop/salt/minion.py#L470
>
> I do not doubt that I have made a mistake in here
>
>
> I see that you are unregistering, closing the socket, and then immediately
> creating a new socket with the same IDENTITY.
>
> I did not see any code setting the LINGER option to drop packets, so if a
> late packet arrives I suppose it could cause that zmq_close() to hang.
>
> Alternately, reusing the IDENTITY before the socket is completely closed
> (an async operation) may be a cause. Any chance you can close the socket
> and then schedule (within your loop) the creation of a replacement socket a
> few milliseconds later?
>
> These are just two guesses based on 5m review of the code. I hope it
> sparks an idea or a new avenue to pursue.
>
> cr
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
> Thanks!
Is there a way to check if the socket is completely closed? will
socket.closed accurately return this data?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/7468fb69/attachment.htm 

From edwinamsler at thinkboxsoftware.com  Mon Jul  9 22:24:21 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Mon, 09 Jul 2012 15:24:21 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water mark
Message-ID: <4FFB3DF5.7010407@thinkboxsoftware.com>

So here I am, publishing messages through ZeroMQ's send() function at 
about 300MB/s, and my network's set to only send at 10MB/s.

This is kind of a big problem because, while I don't care if the clients 
loose data on their end, the server is either using its memory until it 
crashes, or I'm setting its high water mark and loosing about 29 of 
every thirty messages I produce because I don't know that ZeroMQ can't 
keep up.

Ideally, when a HWM condition happened, send() would return false, then 
I'd test EAGAIN so I could decide for myself whether I should drop the 
message, or retry later. With that kind of functionality, I could 
throttle back my producer algorithm so that I exactly meet the demand of 
ZeroMQ instead of overwhelming/starving it out.

I'm willing to do the work if this sort of addition makes sense to the 
rest of the project. I'd rather contribute here instead of forking it 
off in some forgotten repository.

Can/should this be done? Is there someone out there willing to mentor me?

Thanks,

Edwin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/440bc088/attachment.htm 

From thatch45 at gmail.com  Mon Jul  9 22:26:24 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Mon, 9 Jul 2012 14:26:24 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
	<CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
	<B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>
	<CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>
Message-ID: <CAO18cg-s5WFaWQLFbYSTGeg2F-vPtApR8cvc9BhLCfGiMARhww@mail.gmail.com>

On Mon, Jul 9, 2012 at 1:59 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:

>
>
> On Mon, Jul 9, 2012 at 1:43 PM, Chuck Remes <lists at chuckremes.com> wrote:
>
>> On Jul 9, 2012, at 2:17 PM, Thomas S Hatch wrote:
>>
>> Yes, they are, here is the code:
>> https://github.com/saltstack/salt/blob/develop/salt/minion.py#L470
>>
>> I do not doubt that I have made a mistake in here
>>
>>
>> I see that you are unregistering, closing the socket, and then
>> immediately creating a new socket with the same IDENTITY.
>>
>> I did not see any code setting the LINGER option to drop packets, so if a
>> late packet arrives I suppose it could cause that zmq_close() to hang.
>>
>> Alternately, reusing the IDENTITY before the socket is completely closed
>> (an async operation) may be a cause. Any chance you can close the socket
>> and then schedule (within your loop) the creation of a replacement socket a
>> few milliseconds later?
>>
>> These are just two guesses based on 5m review of the code. I hope it
>> sparks an idea or a new avenue to pursue.
>>
>> cr
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>> Thanks!
> Is there a way to check if the socket is completely closed? will
> socket.closed accurately return this data?
>
Actually, I should ask...

The main issue we are seeing in here is that the SUB socket looses
connection from the PUB socket when connecting over unreliable connections
like the internet. Is there a way to detect if the SUB socket is still
connected to the PUB socket?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/78dd3bfb/attachment.htm 

From pelletier.michel at gmail.com  Mon Jul  9 22:28:27 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 9 Jul 2012 13:28:27 -0700
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
	<CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
	<B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>
	<CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>
Message-ID: <CACxu=vLHNhnWUDO3Ydfp5oHrospCWKmT09SU7FZgvsuiuscdOw@mail.gmail.com>


From ian.barber at gmail.com  Mon Jul  9 22:29:19 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Mon, 9 Jul 2012 21:29:19 +0100
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFB3DF5.7010407@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
Message-ID: <CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>

On Mon, Jul 9, 2012 at 9:24 PM, Edwin Amsler <
edwinamsler at thinkboxsoftware.com> wrote:

>
> Ideally, when a HWM condition happened, send() would return false, then
> I'd test EAGAIN so I could decide for myself whether I should drop the
> message, or retry later. With that kind of functionality, I could throttle
> back my producer algorithm so that I exactly meet the demand of ZeroMQ
> instead of overwhelming/starving it out.
>
>
This is pretty much how it works with most socket types - if you use the
ZMQ_NOBLOCK flag, you'll get the EAGAIN if you've hit HWM.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/d30a445c/attachment.htm 

From edwinamsler at thinkboxsoftware.com  Mon Jul  9 22:31:43 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Mon, 09 Jul 2012 15:31:43 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
 mark
In-Reply-To: <CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
Message-ID: <4FFB3FAF.8000809@thinkboxsoftware.com>

Guh? Reeeaally? Through my testing the publisher never errored on send, 
so I didn't test EAGAIN.

A large part of the project is in metaphorical pieces on the floor while 
I redesign something. When I get it back together, I'll test this. Can 
anyone else confirm that that's what happens even on publisher sockets?

On 09/07/2012 3:29 PM, Ian Barber wrote:
>
>
> On Mon, Jul 9, 2012 at 9:24 PM, Edwin Amsler 
> <edwinamsler at thinkboxsoftware.com 
> <mailto:edwinamsler at thinkboxsoftware.com>> wrote:
>
>
>     Ideally, when a HWM condition happened, send() would return false,
>     then I'd test EAGAIN so I could decide for myself whether I should
>     drop the message, or retry later. With that kind of functionality,
>     I could throttle back my producer algorithm so that I exactly meet
>     the demand of ZeroMQ instead of overwhelming/starving it out.
>
>
> This is pretty much how it works with most socket types - if you use 
> the ZMQ_NOBLOCK flag, you'll get the EAGAIN if you've hit HWM.
>
> Ian
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/b66eff22/attachment.htm 

From ian.barber at gmail.com  Mon Jul  9 22:32:28 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Mon, 9 Jul 2012 21:32:28 +0100
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
Message-ID: <CAEYnEwGwvPz70OOB=3hVOqWZqTviq+_TbJ-z-i1iZNBMjBMYLg@mail.gmail.com>

On Mon, Jul 9, 2012 at 9:29 PM, Ian Barber <ian.barber at gmail.com> wrote:

>
>>
> This is pretty much how it works with most socket types - if you use the
> ZMQ_NOBLOCK flag, you'll get the EAGAIN if you've hit HWM.
>
> Ian
>

You can see all of the HWM responses here:
http://api.zeromq.org/2-1:zmq-socket

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/3cfbeb44/attachment.htm 

From ian.barber at gmail.com  Mon Jul  9 22:34:02 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Mon, 9 Jul 2012 21:34:02 +0100
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAO18cg-s5WFaWQLFbYSTGeg2F-vPtApR8cvc9BhLCfGiMARhww@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
	<CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
	<B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>
	<CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>
	<CAO18cg-s5WFaWQLFbYSTGeg2F-vPtApR8cvc9BhLCfGiMARhww@mail.gmail.com>
Message-ID: <CAEYnEwFJ=KO30p_HpMWyN8USAzJQ38tt=j1E86DdXHLszub=JQ@mail.gmail.com>

On Mon, Jul 9, 2012 at 9:26 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:
>
> Actually, I should ask...
>
> The main issue we are seeing in here is that the SUB socket looses
> connection from the PUB socket when connecting over unreliable connections
> like the internet. Is there a way to detect if the SUB socket is still
> connected to the PUB socket?
>

Sort of: Either layer a heartbeat on top, or use the ZMQ_MONITOR stuff in
ZMQ3.1+ to look for disconnect events. I suspect you'll get better accuracy
with the heartbeat (as it'll catch a multitude of sins), but monitor would
easier.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/99515270/attachment.htm 

From pelletier.michel at gmail.com  Mon Jul  9 22:41:43 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 9 Jul 2012 13:41:43 -0700
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFB3DF5.7010407@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
Message-ID: <CACxu=vKYTkpUDN_kOJ4AsarqnsGTe5XsDK60xX4VqaaWahXcEg@mail.gmail.com>

Have you looked at the suicidal snail pattern?

http://zguide.zeromq.org/page:all#Slow-Subscriber-Detection-Suicidal-Snail-Pattern

-Michel

On Mon, Jul 9, 2012 at 1:24 PM, Edwin Amsler
<edwinamsler at thinkboxsoftware.com> wrote:
> So here I am, publishing messages through ZeroMQ's send() function at about
> 300MB/s, and my network's set to only send at 10MB/s.
>
> This is kind of a big problem because, while I don't care if the clients
> loose data on their end, the server is either using its memory until it
> crashes, or I'm setting its high water mark and loosing about 29 of every
> thirty messages I produce because I don't know that ZeroMQ can't keep up.
>
> Ideally, when a HWM condition happened, send() would return false, then I'd
> test EAGAIN so I could decide for myself whether I should drop the message,
> or retry later. With that kind of functionality, I could throttle back my
> producer algorithm so that I exactly meet the demand of ZeroMQ instead of
> overwhelming/starving it out.
>
> I'm willing to do the work if this sort of addition makes sense to the rest
> of the project. I'd rather contribute here instead of forking it off in some
> forgotten repository.
>
> Can/should this be done? Is there someone out there willing to mentor me?
>
> Thanks,
>
> Edwin
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From edwinamsler at thinkboxsoftware.com  Mon Jul  9 22:44:45 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Mon, 09 Jul 2012 15:44:45 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
 mark
In-Reply-To: <CACxu=vKYTkpUDN_kOJ4AsarqnsGTe5XsDK60xX4VqaaWahXcEg@mail.gmail.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CACxu=vKYTkpUDN_kOJ4AsarqnsGTe5XsDK60xX4VqaaWahXcEg@mail.gmail.com>
Message-ID: <4FFB42BD.1030506@thinkboxsoftware.com>

No need. I don't even care about subscribers. It's all about feeding the 
sending state machine efficiently.

On 09/07/2012 3:41 PM, Michel Pelletier wrote:
> Have you looked at the suicidal snail pattern?
>
> http://zguide.zeromq.org/page:all#Slow-Subscriber-Detection-Suicidal-Snail-Pattern
>
> -Michel
>
> On Mon, Jul 9, 2012 at 1:24 PM, Edwin Amsler
> <edwinamsler at thinkboxsoftware.com>  wrote:
>> So here I am, publishing messages through ZeroMQ's send() function at about
>> 300MB/s, and my network's set to only send at 10MB/s.
>>
>> This is kind of a big problem because, while I don't care if the clients
>> loose data on their end, the server is either using its memory until it
>> crashes, or I'm setting its high water mark and loosing about 29 of every
>> thirty messages I produce because I don't know that ZeroMQ can't keep up.
>>
>> Ideally, when a HWM condition happened, send() would return false, then I'd
>> test EAGAIN so I could decide for myself whether I should drop the message,
>> or retry later. With that kind of functionality, I could throttle back my
>> producer algorithm so that I exactly meet the demand of ZeroMQ instead of
>> overwhelming/starving it out.
>>
>> I'm willing to do the work if this sort of addition makes sense to the rest
>> of the project. I'd rather contribute here instead of forking it off in some
>> forgotten repository.
>>
>> Can/should this be done? Is there someone out there willing to mentor me?
>>
>> Thanks,
>>
>> Edwin
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From edwinamsler at thinkboxsoftware.com  Mon Jul  9 22:46:18 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Mon, 09 Jul 2012 15:46:18 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
 mark
In-Reply-To: <CAEYnEwGwvPz70OOB=3hVOqWZqTviq+_TbJ-z-i1iZNBMjBMYLg@mail.gmail.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
	<CAEYnEwGwvPz70OOB=3hVOqWZqTviq+_TbJ-z-i1iZNBMjBMYLg@mail.gmail.com>
Message-ID: <4FFB431A.8030503@thinkboxsoftware.com>

There doesn't seem to be anything about how it signals that the high 
water mark happened. Unless I didn't read it thoroughly enough.

"When a/ZMQ_PUB/socket enters an exceptional state due to having reached 
the high water mark for a/subscriber/, then any messages that would be 
sent to the/subscriber/in question shall instead be dropped until the 
exceptional state ends. The/zmq_send()/function shall never block for 
this socket type."

On 09/07/2012 3:32 PM, Ian Barber wrote:
>
>
> On Mon, Jul 9, 2012 at 9:29 PM, Ian Barber <ian.barber at gmail.com 
> <mailto:ian.barber at gmail.com>> wrote:
>
>
>
>     This is pretty much how it works with most socket types - if you
>     use the ZMQ_NOBLOCK flag, you'll get the EAGAIN if you've hit HWM.
>
>     Ian
>
>
> You can see all of the HWM responses here: 
> http://api.zeromq.org/2-1:zmq-socket
>
> Ian
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/a2026a85/attachment.htm 

From lists at chuckremes.com  Mon Jul  9 22:52:01 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 9 Jul 2012 15:52:01 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFB431A.8030503@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
	<CAEYnEwGwvPz70OOB=3hVOqWZqTviq+_TbJ-z-i1iZNBMjBMYLg@mail.gmail.com>
	<4FFB431A.8030503@thinkboxsoftware.com>
Message-ID: <E5D322B3-326E-4EEB-B828-E5A01793708C@chuckremes.com>

On Jul 9, 2012, at 3:46 PM, Edwin Amsler wrote:

> There doesn't seem to be anything about how it signals that the high water mark happened. Unless I didn't read it thoroughly enough.
> 
> "When a ZMQ_PUB socket enters an exceptional state due to having reached the high water mark for a subscriber, then any messages that would be sent to the subscriber in question shall instead be dropped until the exceptional state ends. The zmq_send()function shall never block for this socket type."

Ed, time for you to write a few lines of code to prove this works (or doesn't) for your use case. You can figure it out a lot faster than asking here now that you've been given some direction.

Let us know how it works out.

cr

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/fe8c025e/attachment.htm 

From edwinamsler at thinkboxsoftware.com  Mon Jul  9 22:54:31 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Mon, 09 Jul 2012 15:54:31 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
 mark
In-Reply-To: <E5D322B3-326E-4EEB-B828-E5A01793708C@chuckremes.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
	<CAEYnEwGwvPz70OOB=3hVOqWZqTviq+_TbJ-z-i1iZNBMjBMYLg@mail.gmail.com>
	<4FFB431A.8030503@thinkboxsoftware.com>
	<E5D322B3-326E-4EEB-B828-E5A01793708C@chuckremes.com>
Message-ID: <4FFB4507.2060704@thinkboxsoftware.com>

I prefer Edwin. I'll make a test case, then report back.

On 09/07/2012 3:52 PM, Chuck Remes wrote:
> On Jul 9, 2012, at 3:46 PM, Edwin Amsler wrote:
>
>> There doesn't seem to be anything about how it signals that the high 
>> water mark happened. Unless I didn't read it thoroughly enough.
>>
>> "When a/ZMQ_PUB/socket enters an exceptional state due to having 
>> reached the high water mark for a/subscriber/, then any messages that 
>> would be sent to the/subscriber/in question shall instead be dropped 
>> until the exceptional state ends. The/zmq_send()/function shall never 
>> block for this socket type."
>
> Ed, time for you to write a few lines of code to prove this works (or 
> doesn't) for your use case. You can figure it out a lot faster than 
> asking here now that you've been given some direction.
>
> Let us know how it works out.
>
> cr
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120709/a936f1c9/attachment.htm 

From pelletier.michel at gmail.com  Mon Jul  9 23:00:28 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 9 Jul 2012 14:00:28 -0700
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFB42BD.1030506@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CACxu=vKYTkpUDN_kOJ4AsarqnsGTe5XsDK60xX4VqaaWahXcEg@mail.gmail.com>
	<4FFB42BD.1030506@thinkboxsoftware.com>
Message-ID: <CACxu=vJeXR1Lj44phNdiQ8=x4BraB+-EqyQFO=P22drGwkQC2A@mail.gmail.com>

On Mon, Jul 9, 2012 at 1:44 PM, Edwin Amsler
<edwinamsler at thinkboxsoftware.com> wrote:
> No need. I don't even care about subscribers. It's all about feeding the
> sending state machine efficiently.

The SSP is about never getting to the point where the PUB socket drops
messages, because your subscribes never go slow enough to fill the
queue.  If you don't care about subscribers, then it sounds like the
right pattern for you.

-Michel

>
> On 09/07/2012 3:41 PM, Michel Pelletier wrote:
>> Have you looked at the suicidal snail pattern?
>>
>> http://zguide.zeromq.org/page:all#Slow-Subscriber-Detection-Suicidal-Snail-Pattern
>>
>> -Michel
>>
>> On Mon, Jul 9, 2012 at 1:24 PM, Edwin Amsler
>> <edwinamsler at thinkboxsoftware.com>  wrote:
>>> So here I am, publishing messages through ZeroMQ's send() function at about
>>> 300MB/s, and my network's set to only send at 10MB/s.
>>>
>>> This is kind of a big problem because, while I don't care if the clients
>>> loose data on their end, the server is either using its memory until it
>>> crashes, or I'm setting its high water mark and loosing about 29 of every
>>> thirty messages I produce because I don't know that ZeroMQ can't keep up.
>>>
>>> Ideally, when a HWM condition happened, send() would return false, then I'd
>>> test EAGAIN so I could decide for myself whether I should drop the message,
>>> or retry later. With that kind of functionality, I could throttle back my
>>> producer algorithm so that I exactly meet the demand of ZeroMQ instead of
>>> overwhelming/starving it out.
>>>
>>> I'm willing to do the work if this sort of addition makes sense to the rest
>>> of the project. I'd rather contribute here instead of forking it off in some
>>> forgotten repository.
>>>
>>> Can/should this be done? Is there someone out there willing to mentor me?
>>>
>>> Thanks,
>>>
>>> Edwin
>>>
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From lists at chuckremes.com  Mon Jul  9 23:08:24 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 9 Jul 2012 16:08:24 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFB4507.2060704@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAEYnEwG0AhvTSBKnFzaaE9MoGs2WWQ1Sjy=qSr7BWOMw8RMo1g@mail.gmail.com>
	<CAEYnEwGwvPz70OOB=3hVOqWZqTviq+_TbJ-z-i1iZNBMjBMYLg@mail.gmail.com>
	<4FFB431A.8030503@thinkboxsoftware.com>
	<E5D322B3-326E-4EEB-B828-E5A01793708C@chuckremes.com>
	<4FFB4507.2060704@thinkboxsoftware.com>
Message-ID: <40076981-40A0-4497-BD1D-5AFEDA0E94A3@chuckremes.com>

On Jul 9, 2012, at 3:54 PM, Edwin Amsler wrote:

> I prefer Edwin. I'll make a test case, then report back.

Noted for future posts. We look forward to hearing from you.

cr

From paul at colomiets.name  Tue Jul 10 08:59:53 2012
From: paul at colomiets.name (Paul Colomiets)
Date: Tue, 10 Jul 2012 09:59:53 +0300
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFB3DF5.7010407@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
Message-ID: <CAA0gF6r79ZgA9ma7xd3YL-spHLEz3FMdxQQqabGxMFTAp7r=-A@mail.gmail.com>

Hi Edwin,

On Mon, Jul 9, 2012 at 11:24 PM, Edwin Amsler
<edwinamsler at thinkboxsoftware.com> wrote:
> So here I am, publishing messages through ZeroMQ's send() function at about
> 300MB/s, and my network's set to only send at 10MB/s.
>
> This is kind of a big problem because, while I don't care if the clients
> loose data on their end, the server is either using its memory until it
> crashes, or I'm setting its high water mark and loosing about 29 of every
> thirty messages I produce because I don't know that ZeroMQ can't keep up.
>
> Ideally, when a HWM condition happened, send() would return false, then I'd
> test EAGAIN so I could decide for myself whether I should drop the message,
> or retry later. With that kind of functionality, I could throttle back my
> producer algorithm so that I exactly meet the demand of ZeroMQ instead of
> overwhelming/starving it out.
>
> I'm willing to do the work if this sort of addition makes sense to the rest
> of the project. I'd rather contribute here instead of forking it off in some
> forgotten repository.
>
> Can/should this be done? Is there someone out there willing to mentor me?
>

The behavior is intentional for pub/sub sockets. If you'd have only
one subscriber you could use push/pull. Push sockets block when reach
high water mark, so are Req sockets.

The pub/sub sockets can't reliably block in general case because there
could be multiple subscribers, only one of which reaches high water
mark. Partially this comes from implementation: when you do zmq_send()
zeromq starts to push message to the pipe for each connected
subscribers, if one of the pipes full, there is no way to rollback
messages already put into pipes, if it would it's not clear whether
single slow subscriber should stop the publisher.  Also pgm sockets
can't have backpressure AFAIU.

You can try to implement the behavior as a socket option (it can't be
default behavior), but you have to be aware of the problems above (And
I don't know if core developers are willing to accept the patch).

-- 
Paul

From edwinamsler at thinkboxsoftware.com  Tue Jul 10 17:31:08 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Tue, 10 Jul 2012 10:31:08 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
 mark
In-Reply-To: <CAA0gF6r79ZgA9ma7xd3YL-spHLEz3FMdxQQqabGxMFTAp7r=-A@mail.gmail.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAA0gF6r79ZgA9ma7xd3YL-spHLEz3FMdxQQqabGxMFTAp7r=-A@mail.gmail.com>
Message-ID: <4FFC4ABC.4070600@thinkboxsoftware.com>

On 10/07/2012 1:59 AM, Paul Colomiets wrote:
> Hi Edwin,
>
> On Mon, Jul 9, 2012 at 11:24 PM, Edwin Amsler
> <edwinamsler at thinkboxsoftware.com>  wrote:
>> So here I am, publishing messages through ZeroMQ's send() function at about
>> 300MB/s, and my network's set to only send at 10MB/s.
>>
>> This is kind of a big problem because, while I don't care if the clients
>> loose data on their end, the server is either using its memory until it
>> crashes, or I'm setting its high water mark and loosing about 29 of every
>> thirty messages I produce because I don't know that ZeroMQ can't keep up.
>>
>> Ideally, when a HWM condition happened, send() would return false, then I'd
>> test EAGAIN so I could decide for myself whether I should drop the message,
>> or retry later. With that kind of functionality, I could throttle back my
>> producer algorithm so that I exactly meet the demand of ZeroMQ instead of
>> overwhelming/starving it out.
>>
>> I'm willing to do the work if this sort of addition makes sense to the rest
>> of the project. I'd rather contribute here instead of forking it off in some
>> forgotten repository.
>>
>> Can/should this be done? Is there someone out there willing to mentor me?
>>
> The behavior is intentional for pub/sub sockets. If you'd have only
> one subscriber you could use push/pull. Push sockets block when reach
> high water mark, so are Req sockets.
>
> The pub/sub sockets can't reliably block in general case because there
> could be multiple subscribers, only one of which reaches high water
> mark. Partially this comes from implementation: when you do zmq_send()
I never want them to block. That'd kind of break my program's state 
machine. What I'd like, is to know when the high water mark was reached 
with a send() error, and and EAGAIN errno so that I can /choose/ whether 
or not I'd like to throw away the message. Right now, the message is 
just silently cast away further down the stack as far as I can tell.
> zeromq starts to push message to the pipe for each connected
> subscribers, if one of the pipes full, there is no way to rollback
> messages already put into pipes, if it would it's not clear whether
> single slow subscriber should stop the publisher.  Also pgm sockets
> can't have backpressure AFAIU.
Is there a way to detect if a pipe is full, and return a send() error 
then? As I said, I feel that pub sockets should never block (that's why 
I changed from using OpenPGM alone to ZeroMQ actually, though I've found 
yet more reasons to stay).
> You can try to implement the behavior as a socket option (it can't be
> default behavior), but you have to be aware of the problems above (And
> I don't know if core developers are willing to accept the patch).
That's more what I'm worried about. I'm quite confident there's no way 
to handle this yet, and I feel that I'm not the only one who'd like to 
have some sense of how fast a pub socket is truthfully sending data. The 
implementation on the API side should be solid since the send() error, 
EAGAIN method is how all ZMQ_NOBLOCK sockets work already. It's just 
that send() doesn't error when a PUB socket starts throwing away data (I 
know that's the case for sure on Windows at least).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/f8595faf/attachment.htm 

From XavierMillieret at Eaton.com  Tue Jul 10 17:48:10 2012
From: XavierMillieret at Eaton.com (XavierMillieret at Eaton.com)
Date: Tue, 10 Jul 2012 15:48:10 +0000
Subject: [zeromq-dev] Jzmq
Message-ID: <B8C3F62415206148B914194A55363BB6C1D4@SIMTCSMB06.napa.ad.etn.com>

Hi,

I try to use the java binding for zeromq on win seven 64:

Configuration:
I downloaded the following libraries:

http://miru.hk/archive/ZeroMQ-3.1.1beta~miru1.1-win64.exe  ->
libzmq-v100-mt.dll

and http://miru.hk/archive/JZMQ-2.1.10~miru1.2-win64.exe  ->
JZMQ 2.1.10\ jzmq.dll
JZMQ 2.1.10\zmq.jar

The problem is In my java code, when I try to implement a pub/sub sample, I am "wake up" on the good topic, but in my receive buffer (String),  it's empty !
I uses the  following  virtual machine:
java version "1.7.0_01"
Java(TM) SE Runtime Environment (build 1.7.0_01-b08)
Java HotSpot(TM) 64-Bit Server VM (build 21.1-b02, mixed mode)

Somebody could help me for my project, please.

Thanks you by advance.


Xavier Millieret




________________________________

________________________________


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/d51d6e2c/attachment.htm 

From gdiethelm at dcv.cl  Tue Jul 10 17:52:19 2012
From: gdiethelm at dcv.cl (gonzalo diethelm)
Date: Tue, 10 Jul 2012 15:52:19 +0000
Subject: [zeromq-dev] Jzmq
In-Reply-To: <B8C3F62415206148B914194A55363BB6C1D4@SIMTCSMB06.napa.ad.etn.com>
References: <B8C3F62415206148B914194A55363BB6C1D4@SIMTCSMB06.napa.ad.etn.com>
Message-ID: <6D436FBF9A7B114392D84BF6832C5CA704978B@aowmail1.dcv.net>

If you don?t subscribe to a specific pattern (which could be empty), you will not get any messages. HTH.

--
Gonzalo Diethelm
DCV Chile

From: zeromq-dev-bounces at lists.zeromq.org [mailto:zeromq-dev-bounces at lists.zeromq.org] On Behalf Of XavierMillieret at Eaton.com
Sent: Tuesday, July 10, 2012 11:48 AM
To: zeromq-dev at lists.zeromq.org
Subject: [zeromq-dev] Jzmq

Hi,

I try to use the java binding for zeromq on win seven 64:

Configuration:
I downloaded the following libraries:

http://miru.hk/archive/ZeroMQ-3.1.1beta~miru1.1-win64.exe  ->
libzmq-v100-mt.dll

and http://miru.hk/archive/JZMQ-2.1.10~miru1.2-win64.exe  ->
JZMQ 2.1.10\ jzmq.dll
JZMQ 2.1.10\zmq.jar

The problem is In my java code, when I try to implement a pub/sub sample, I am ?wake up? on the good topic, but in my receive buffer (String),  it?s empty !
I uses the  following  virtual machine:
java version "1.7.0_01"
Java(TM) SE Runtime Environment (build 1.7.0_01-b08)
Java HotSpot(TM) 64-Bit Server VM (build 21.1-b02, mixed mode)

Somebody could help me for my project, please.

Thanks you by advance.


Xavier Millieret




________________________________
________________________________



----------------------------------------- 
Declaraci?n de confidencialidad: Este Mensaje esta destinado para
el uso de la o las personas o entidades a quien ha sido dirigido y
puede contener informaci?n reservada y confidencial que no puede
ser divulgada, difundida, ni aprovechada en forma alguna. El uso no
autorizado de la informaci?n contenida en este correo podr? ser
sancionado de conformidad con la ley chilena. 
Si usted ha recibido este correo electr?nico por error, le pedimos
eliminarlo junto con los archivos adjuntos y avisar inmediatamente
al remitente, respondiendo este mensaje. 

"Before printing this e-mail think if is really necesary".
Disclosure: This Message is to be used by the individual,
individuals or entities that it is addressed to and may include
private and confidential information that may not be disclosed,
made public nor used in any way at all. Unauthorized use of the
information in this electronic mail message may be subject to the
penalties set forth by Chilean law. 
If you have received this electronic mail message in error, we ask
you to destroy the message and its attached file(s) and to
immediately notify the sender by answering this message. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/1534a253/attachment.htm 

From lists at chuckremes.com  Tue Jul 10 18:00:07 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Tue, 10 Jul 2012 11:00:07 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
	mark
In-Reply-To: <4FFC4ABC.4070600@thinkboxsoftware.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAA0gF6r79ZgA9ma7xd3YL-spHLEz3FMdxQQqabGxMFTAp7r=-A@mail.gmail.com>
	<4FFC4ABC.4070600@thinkboxsoftware.com>
Message-ID: <0E1ED34A-B052-41E9-AB5C-260FDA31E6BF@chuckremes.com>


On Jul 10, 2012, at 10:31 AM, Edwin Amsler wrote:

> On 10/07/2012 1:59 AM, Paul Colomiets wrote:
>> 
>> Hi Edwin,
>> 
>> The behavior is intentional for pub/sub sockets. If you'd have only
>> one subscriber you could use push/pull. Push sockets block when reach
>> high water mark, so are Req sockets.
>> 
>> The pub/sub sockets can't reliably block in general case because there
>> could be multiple subscribers, only one of which reaches high water
>> mark. Partially this comes from implementation: when you do zmq_send()
> I never want them to block. That'd kind of break my program's state machine. What I'd like, is to know when the high water mark was reached with a send() error, and and EAGAIN errno so that I can choose whether or not I'd like to throw away the message. Right now, the message is just silently cast away further down the stack as far as     I can tell.

I need to stress a particular detail here. The socket has a separate HWM for *each* connected subscriber because each one has its own outgoing queue. If you were to receive EAGAIN (or an error return code) from zmq_send() when a subset of subscribers hit their HWM, your published message would still be received by subscribers that have not fallen behind. It does *not* drop the message for all subscribers, just those that have entered the HWM exceptional state.

I don't think you can (even with a patch specific to your use case) rely on this behavior to achieve your goal.

>> zeromq starts to push message to the pipe for each connected
>> subscribers, if one of the pipes full, there is no way to rollback
>> messages already put into pipes, if it would it's not clear whether
>> single slow subscriber should stop the publisher.  Also pgm sockets
>> can't have backpressure AFAIU.
> Is there a way to detect if a pipe is full, and return a send() error then? As I said, I feel that pub sockets should never block (that's why I changed from using OpenPGM alone to ZeroMQ actually, though I've found yet more reasons to stay). 

Not at this time. However, there has been some work done by others to let developers receive notification for specific internal library events. I'm certain it would be possible to add this type of event as a notification.

Also, PUB sockets never block.

>> You can try to implement the behavior as a socket option (it can't be
>> default behavior), but you have to be aware of the problems above (And
>> I don't know if core developers are willing to accept the patch).
> That's more what I'm worried about. I'm quite confident there's no way to handle this yet, and I feel that I'm not the only one who'd like to have some sense of how fast a pub socket is truthfully sending data. The implementation on the API side should be solid since the send() error, EAGAIN method is how all ZMQ_NOBLOCK sockets work already. It's just that send() doesn't error when a PUB socket starts throwing away data (I know that's the case for sure on Windows at least).

What we really need is a better mechanism for handling this type of situation than HWM. One particular thought is described here:

http://unprotocols.org/blog:15

This can be implemented today using DEALER/ROUTER sockets.

You may also be interested in the Suicidal Snail pattern described in the guide.

http://zguide.zeromq.org/page:all#Slow-Subscriber-Detection-Suicidal-Snail-Pattern

If you haven't read the guide yet, you should do so before getting too much deeper into your problem space. It will save you a lot of time.

cr


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/6baf8de1/attachment.htm 

From edwinamsler at thinkboxsoftware.com  Tue Jul 10 18:18:35 2012
From: edwinamsler at thinkboxsoftware.com (Edwin Amsler)
Date: Tue, 10 Jul 2012 11:18:35 -0500
Subject: [zeromq-dev] Notify send()er that they've hit the high water
 mark
In-Reply-To: <0E1ED34A-B052-41E9-AB5C-260FDA31E6BF@chuckremes.com>
References: <4FFB3DF5.7010407@thinkboxsoftware.com>
	<CAA0gF6r79ZgA9ma7xd3YL-spHLEz3FMdxQQqabGxMFTAp7r=-A@mail.gmail.com>
	<4FFC4ABC.4070600@thinkboxsoftware.com>
	<0E1ED34A-B052-41E9-AB5C-260FDA31E6BF@chuckremes.com>
Message-ID: <4FFC55DB.7020403@thinkboxsoftware.com>

On 10/07/2012 11:00 AM, Chuck Remes wrote:
>
> On Jul 10, 2012, at 10:31 AM, Edwin Amsler wrote:
>
>> On 10/07/2012 1:59 AM, Paul Colomiets wrote:
>>> Hi Edwin,
>>>
>>> The behavior is intentional for pub/sub sockets. If you'd have only
>>> one subscriber you could use push/pull. Push sockets block when reach
>>> high water mark, so are Req sockets.
>>>
>>> The pub/sub sockets can't reliably block in general case because there
>>> could be multiple subscribers, only one of which reaches high water
>>> mark. Partially this comes from implementation: when you do zmq_send()
>> I never want them to block. That'd kind of break my program's state 
>> machine. What I'd like, is to know when the high water mark was 
>> reached with a send() error, and and EAGAIN errno so that I can 
>> /choose/ whether or not I'd like to throw away the message. Right 
>> now, the message is just silently cast away further down the stack as 
>> far as I can tell.
>
> I need to stress a particular detail here. The socket has a separate 
> HWM for *each* connected subscriber because each one has its own 
> outgoing queue. If you were to receive EAGAIN (or an error return 
> code) from zmq_send() when a subset of subscribers hit their HWM, your 
> published message would still be received by subscribers that have not 
> fallen behind. It does *not* drop the message for all subscribers, 
> just those that have entered the HWM exceptional state.
>
> I don't think you can (even with a patch specific to your use case) 
> rely on this behavior to achieve your goal.
Well, that's depressing. I'm using "epgm://" here, and I know that 
OpenPGM itself has no concept of connected clients. Are there still 
individual queues for clients in this case?
>
>>> zeromq starts to push message to the pipe for each connected
>>> subscribers, if one of the pipes full, there is no way to rollback
>>> messages already put into pipes, if it would it's not clear whether
>>> single slow subscriber should stop the publisher.  Also pgm sockets
>>> can't have backpressure AFAIU.
>> Is there a way to detect if a pipe is full, and return a send() error 
>> then? As I said, I feel that pub sockets should never block (that's 
>> why I changed from using OpenPGM alone to ZeroMQ actually, though 
>> I've found yet more reasons to stay).
>
> Not at this time. However, there has been some work done by others to 
> let developers receive notification for specific internal library 
> events. I'm certain it would be possible to add this type of event as 
> a notification.
>
> Also, PUB sockets never block.
>
>>> You can try to implement the behavior as a socket option (it can't be
>>> default behavior), but you have to be aware of the problems above (And
>>> I don't know if core developers are willing to accept the patch).
>> That's more what I'm worried about. I'm quite confident there's no 
>> way to handle this yet, and I feel that I'm not the only one who'd 
>> like to have some sense of how fast a pub socket is truthfully 
>> sending data. The implementation on the API side should be solid 
>> since the send() error, EAGAIN method is how all ZMQ_NOBLOCK sockets 
>> work already. It's just that send() doesn't error when a PUB socket 
>> starts throwing away data (I know that's the case for sure on Windows 
>> at least).
>
> What we really need is a better mechanism for handling this type of 
> situation than HWM. One particular thought is described here:
>
> http://unprotocols.org/blog:15
>
> This can be implemented today using DEALER/ROUTER sockets.
>
> You may also be interested in the Suicidal Snail pattern described in 
> the guide.
>
> http://zguide.zeromq.org/page:all#Slow-Subscriber-Detection-Suicidal-Snail-Pattern
>
> If you haven't read the guide yet, you should do so before getting too 
> much deeper into your problem space. It will save you a lot of time.
Good advice. I made it about three chapters in. I really need to go 
digest it though.
>
> cr
>
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/0f89fd7f/attachment.htm 

From jhawk28 at gmail.com  Wed Jul 11 00:02:29 2012
From: jhawk28 at gmail.com (Joshua Foster)
Date: Tue, 10 Jul 2012 18:02:29 -0400
Subject: [zeromq-dev] Jzmq
In-Reply-To: <B8C3F62415206148B914194A55363BB6C1D4@SIMTCSMB06.napa.ad.etn.com>
References: <B8C3F62415206148B914194A55363BB6C1D4@SIMTCSMB06.napa.ad.etn.com>
Message-ID: <4FFCA675.4030602@gmail.com>

If you don't do a socket.recv(0), then it will return null if there is 
nothing in the queue.

Joshua

> XavierMillieret at Eaton.com <mailto:XavierMillieret at Eaton.com>
> July 10, 2012 11:48 AM
>
> Hi,
>
> I try to use the java binding for zeromq on win seven 64:
>
> Configuration:
>
> I downloaded the following libraries:
>
> /http://miru.hk/archive/ZeroMQ-3.1.1beta~miru1.1-win64.exe 
> <http://miru.hk/archive/ZeroMQ-3.1.1beta%7Emiru1.1-win64.exe>//  ->/
>
> libzmq-v100-mt.dll
>
> and /http://miru.hk/archive/JZMQ-2.1.10~miru1.2-win64.exe 
> <http://miru.hk/archive/JZMQ-2.1.10%7Emiru1.2-win64.exe>//  ->/
>
> JZMQ 2.1.10\ jzmq.dll
>
> JZMQ 2.1.10\zmq.jar
>
> The problem is In my java code, when I try to implement a pub/sub 
> sample, I am ?wake up? on the good topic, but in my receive buffer 
> (String),  it?s empty !
>
> I uses the  following  virtual machine:
>
> java version "1.7.0_01"
>
> Java(TM) SE Runtime Environment (build 1.7.0_01-b08)
>
> Java HotSpot(TM) 64-Bit Server VM (build 21.1-b02, mixed mode)
>
> Somebody could help me for my project, please.
>
> Thanks you by advance.
>
> *Xavier Millieret*
>
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/e841623e/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: compose-unknown-contact.jpg
Type: image/jpeg
Size: 770 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/e841623e/attachment.jpg 

From Sharon.Ben-Asher at avg.com  Wed Jul 11 12:32:49 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Wed, 11 Jul 2012 10:32:49 +0000
Subject: [zeromq-dev] Error: java.lang.UnsatisfiedLinkError
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FF437@ex10czdmb01>

Hello again,

I am now trying to build the server side (finished the build of the client on Android, thanks)
I am testing my Java binding on RedHat i686.

My command is as follows
java -Djava.library.path=/root/zeromq-3.2.0/output/lib -classpath .:/root/zeromq-3.2.0/output/share/java:./local_lat Test

Getting the following error:

Exception in thread "Thread-2" java.lang.NoClassDefFoundError: Could not initialize class org.zeromq.ZMQ
      at org.zeromq.ZContext.createSocket(ZContext.java:89)
      at Test$Client.run(Test.java:79)
      at java.lang.Thread.run(Thread.java:662)
Exception in thread "Thread-1" java.lang.NoClassDefFoundError: Could not initialize class org.zeromq.ZMQ
      at org.zeromq.ZContext.createSocket(ZContext.java:89)
      at Test$Worker.run(Test.java:59)
      at java.lang.Thread.run(Thread.java:662)
Exception in thread "Thread-0" java.lang.UnsatisfiedLinkError: /root/zeromq-3.2.0/output/lib/libjzmq.so.0.0.0: libzmq.so.3: cannot open shared object file: No such file or directory
      at java.lang.ClassLoader$NativeLibrary.load(Native Method)
      at java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1807)
      at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1732)
      at java.lang.Runtime.loadLibrary0(Runtime.java:823)
      at java.lang.System.loadLibrary(System.java:1028)
      at org.zeromq.ZMQ.<clinit>(ZMQ.java:34)
      at org.zeromq.ZContext.createSocket(ZContext.java:89)
      at Test$Broker.run(Test.java:19)
      at java.lang.Thread.run(Thread.java:662)

so I am guessing libjzmq.so.0.0.0 can't find libzmq.so.3 ?
however the file is present at /usr/local/lib and I did ldconfig

thanks,

Sharon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120711/16e41399/attachment.htm 

From Sharon.Ben-Asher at avg.com  Wed Jul 11 12:39:01 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Wed, 11 Jul 2012 10:39:01 +0000
Subject: [zeromq-dev] Error: java.lang.UnsatisfiedLinkError
Message-ID: <671E27F33F2F25499F78A4AFA494E63B2E7FF449@ex10czdmb01>

Ok, I got it fixed by following the advice here
https://github.com/zeromq/jzmq/issues/43 (adding the required path to LD_LIBRARY_PATH

thanks!

Sharon

From: Sharon Ben-Asher
Sent: Wednesday, July 11, 2012 1:33 PM
To: 'zeromq-dev at lists.zeromq.org'
Subject: Error: java.lang.UnsatisfiedLinkError

Hello again,

I am now trying to build the server side (finished the build of the client on Android, thanks)
I am testing my Java binding on RedHat i686.

My command is as follows
java -Djava.library.path=/root/zeromq-3.2.0/output/lib -classpath .:/root/zeromq-3.2.0/output/share/java:./local_lat Test

Getting the following error:

Exception in thread "Thread-2" java.lang.NoClassDefFoundError: Could not initialize class org.zeromq.ZMQ
      at org.zeromq.ZContext.createSocket(ZContext.java:89)
      at Test$Client.run(Test.java:79)
      at java.lang.Thread.run(Thread.java:662)
Exception in thread "Thread-1" java.lang.NoClassDefFoundError: Could not initialize class org.zeromq.ZMQ
      at org.zeromq.ZContext.createSocket(ZContext.java:89)
      at Test$Worker.run(Test.java:59)
      at java.lang.Thread.run(Thread.java:662)
Exception in thread "Thread-0" java.lang.UnsatisfiedLinkError: /root/zeromq-3.2.0/output/lib/libjzmq.so.0.0.0: libzmq.so.3: cannot open shared object file: No such file or directory
      at java.lang.ClassLoader$NativeLibrary.load(Native Method)
      at java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1807)
      at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1732)
      at java.lang.Runtime.loadLibrary0(Runtime.java:823)
      at java.lang.System.loadLibrary(System.java:1028)
      at org.zeromq.ZMQ.<clinit>(ZMQ.java:34)
      at org.zeromq.ZContext.createSocket(ZContext.java:89)
      at Test$Broker.run(Test.java:19)
      at java.lang.Thread.run(Thread.java:662)

so I am guessing libjzmq.so.0.0.0 can't find libzmq.so.3 ?
however the file is present at /usr/local/lib and I did ldconfig

thanks,

Sharon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120711/2f993527/attachment.htm 

From XavierMillieret at Eaton.com  Tue Jul 10 17:41:14 2012
From: XavierMillieret at Eaton.com (XavierMillieret at Eaton.com)
Date: Tue, 10 Jul 2012 15:41:14 +0000
Subject: [zeromq-dev] jzmq
Message-ID: <B8C3F62415206148B914194A55363BB6C1B6@SIMTCSMB06.napa.ad.etn.com>

Hi,

I try to use the java binding for zeromq on win seven 64:

Configuration:
I downloaded the following libraries:

http://miru.hk/archive/ZeroMQ-3.1.1beta~miru1.1-win64.exe  ->
libzmq-v100-mt.dll

and http://miru.hk/archive/JZMQ-2.1.10~miru1.2-win64.exe  ->
JZMQ 2.1.10\ jzmq.dll
JZMQ 2.1.10\zmq.jar
The problem is In my java code, when I try to implement a pub/sub sample, I am "wake up" on the good topic, but in my receive buffer (String),  it's empty !
I uses the  following  virtual machine:
java version "1.7.0_01"
Java(TM) SE Runtime Environment (build 1.7.0_01-b08)
Java HotSpot(TM) 64-Bit Server VM (build 21.1-b02, mixed mode)

Somebody could help me for my project, please.

Thanks you by advance.


Xavier Millieret




________________________________

________________________________


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120710/b798ece1/attachment.htm 

From pelletier.michel at gmail.com  Wed Jul 11 17:43:37 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Wed, 11 Jul 2012 08:43:37 -0700
Subject: [zeromq-dev] jzmq
In-Reply-To: <B8C3F62415206148B914194A55363BB6C1B6@SIMTCSMB06.napa.ad.etn.com>
References: <B8C3F62415206148B914194A55363BB6C1B6@SIMTCSMB06.napa.ad.etn.com>
Message-ID: <CACxu=vKmbE=APsTQUBJHCNtgSJ5erhYb=fkkb5urRwNZUrY9fg@mail.gmail.com>

If you're not getting an error it's very likely a problem with your
code.  Please put a _minimal_ example of your code that exhibits your
problem in a pastebin somewhere and send us the link.

-Michel

On Tue, Jul 10, 2012 at 8:41 AM,  <XavierMillieret at eaton.com> wrote:
> Hi,
>
>
>
> I try to use the java binding for zeromq on win seven 64:
>
>
>
> Configuration:
>
> I downloaded the following libraries:
>
> http://miru.hk/archive/ZeroMQ-3.1.1beta~miru1.1-win64.exe  ->
>
> libzmq-v100-mt.dll
>
> and http://miru.hk/archive/JZMQ-2.1.10~miru1.2-win64.exe  ->
>
> JZMQ 2.1.10\ jzmq.dll
>
> JZMQ 2.1.10\zmq.jar
>
> The problem is In my java code, when I try to implement a pub/sub sample, I
> am ?wake up? on the good topic, but in my receive buffer (String),  it?s
> empty !
>
> I uses the  following  virtual machine:
>
> java version "1.7.0_01"
>
> Java(TM) SE Runtime Environment (build 1.7.0_01-b08)
>
> Java HotSpot(TM) 64-Bit Server VM (build 21.1-b02, mixed mode)
>
>
>
> Somebody could help me for my project, please.
>
>
>
> Thanks you by advance.
>
>
>
>
>
> Xavier Millieret
>
>
>
>
>
> ________________________________
>
> ________________________________
>
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From midekk at gmail.com  Wed Jul 11 21:04:06 2012
From: midekk at gmail.com (Midek)
Date: Wed, 11 Jul 2012 21:04:06 +0200
Subject: [zeromq-dev] web based client without flash
Message-ID: <CAGr6vh01Ty=-k38rQWmSGLmg1qaqR-Re07v35X4Awz3D8+PXzQ@mail.gmail.com>

Hello Everybody,

we would like to use 0MQ for new projects. It should server client
architecture for local and remote connection.

The only problem is that one of the clients should be web based,
implemented in javascript later maybe in HTML5. We don't like that
javascript binding use flash.

Is there some another possibility or workaround how to implement web
based client using 0MQ without using flash?

Thank you,

Best regards,

Miro

From ian.barber at gmail.com  Wed Jul 11 21:36:15 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Wed, 11 Jul 2012 20:36:15 +0100
Subject: [zeromq-dev] web based client without flash
In-Reply-To: <CAGr6vh01Ty=-k38rQWmSGLmg1qaqR-Re07v35X4Awz3D8+PXzQ@mail.gmail.com>
References: <CAGr6vh01Ty=-k38rQWmSGLmg1qaqR-Re07v35X4Awz3D8+PXzQ@mail.gmail.com>
Message-ID: <CAEYnEwG-aqVsPEifn5WWU3JegV1MBmJEo9_xkw=geLNtFWk5jg@mail.gmail.com>

On Wed, Jul 11, 2012 at 8:04 PM, Midek <midekk at gmail.com> wrote:

> Hello Everybody,
>
> we would like to use 0MQ for new projects. It should server client
> architecture for local and remote connection.
>
> The only problem is that one of the clients should be web based,
> implemented in javascript later maybe in HTML5. We don't like that
> javascript binding use flash.


Your best bet is likely to use a bridge application - Mongrel2, ZeroGW,
NullMQ, or similar. Without an extension or flash you're not going to be
able to open arbitrary sockets from the browser, which limits your options
somewhat.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120711/4203d6a2/attachment.htm 

From lists at chuckremes.com  Thu Jul 12 00:28:32 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Wed, 11 Jul 2012 17:28:32 -0500
Subject: [zeromq-dev] daily build cluster question
In-Reply-To: <96219A52-FDEA-4377-9ED4-6C17B00C0E6B@chuckremes.com>
References: <96219A52-FDEA-4377-9ED4-6C17B00C0E6B@chuckremes.com>
Message-ID: <DFFB58FE-CA87-4411-BAE9-FBF4BF0805B8@chuckremes.com>

Who is in charge of the daily build cluster? I'd like a login to the debian linux image if that is at all possible. Thanks.

cr

On Jul 2, 2012, at 1:48 PM, Chuck Remes wrote:

> I see on the #zeromq.build channel that the ffi-rzmq gem has failures on Debian linux. I'm the maintainer of that gem and would like to fix those failures. On my OSX box *and* my Archlinux box, the specs all pass. What version of Debian Linux (3, 4, 5 or 6) should I install to match the build cluster so I can reproduce this problem?
> 
> cr
> 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From ian.barber at gmail.com  Thu Jul 12 00:49:55 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Wed, 11 Jul 2012 23:49:55 +0100
Subject: [zeromq-dev] daily build cluster question
In-Reply-To: <DFFB58FE-CA87-4411-BAE9-FBF4BF0805B8@chuckremes.com>
References: <96219A52-FDEA-4377-9ED4-6C17B00C0E6B@chuckremes.com>
	<DFFB58FE-CA87-4411-BAE9-FBF4BF0805B8@chuckremes.com>
Message-ID: <CAEYnEwHS2wfWCMmTwQqattAs3zdr4o5cDKsjjFGa1rOk5Kjdrg@mail.gmail.com>

On Wed, Jul 11, 2012 at 11:28 PM, Chuck Remes <lists at chuckremes.com> wrote:

> Who is in charge of the daily build cluster? I'd like a login to the
> debian linux image if that is at all possible. Thanks.
>
> cr


Mikko.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120711/480dd7cc/attachment.htm 

From segg2 at videotron.ca  Thu Jul 12 06:42:23 2012
From: segg2 at videotron.ca (Gilles J. Seguin)
Date: Thu, 12 Jul 2012 00:42:23 -0400
Subject: [zeromq-dev] [czmq] src/zsocket.c,
	this looks like unused variable to me
Message-ID: <1342068143.17421.6.camel@laure.segg2.videotron.ca>

fedora gcc-4.7.0-5.fc17.x86_64
$ gcc -I../include -Werror -Wno-unused -Wall -c zsocket.c -o zsocket.o
zsocket.c: In function ?zsocket_poll?:
zsocket.c:152:9: error: unused variable ?rc? [-Werror=unused-variable]
cc1: all warnings being treated as errors

src/zsocket.c
@@ -149,7 +149,7 @@ Bool
 zsocket_poll (void *socket, int msecs)
 {
     zmq_pollitem_t items [] = { { socket, 0, ZMQ_POLLIN, 0 } };
-    int rc = zmq_poll (items, 1, msecs);
+    /*int rc =*/ zmq_poll (items, 1, msecs);
     return (items [0].revents & ZMQ_POLLIN) != 0;
 }
 
curious thing, -Wall option just before the -c option





From g.passault at gmail.com  Thu Jul 12 15:43:30 2012
From: g.passault at gmail.com (=?ISO-8859-1?Q?Gr=E9goire_Passault?=)
Date: Thu, 12 Jul 2012 15:43:30 +0200
Subject: [zeromq-dev] Broadcasting with a Router ?
Message-ID: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>

Hi,

Since I'm new to ZeroMQ I really appologize if my question is stupid, but
can we use a Router to broadcast messages to all clients ?

I mean, I don't see any technical reason it couldn't, and I know I also
could create Request/Req and Pub/Sub sockets for instance, but I don't want
to create two of them if one can do the job.

Best regards,
Gr?goire Passaukt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/ad790370/attachment.htm 

From diffuser78 at gmail.com  Thu Jul 12 20:23:10 2012
From: diffuser78 at gmail.com (Diffuser78)
Date: Thu, 12 Jul 2012 11:23:10 -0700
Subject: [zeromq-dev] Quick question on usage of patterns
Message-ID: <CAKni0ed721gwACAzS16W_fakGLx1A2oSBDZAJdXc5H--MfrXiA@mail.gmail.com>

Hello All,

I need some suggestion on my design and ways in which it can be done using
ZMQ sockets.


I have an app  App1 on Box1. It sends a message M1 (128K) to an app App2 on
Box2. App2 will do some processing on M1 before sending an ACK back. In my
application, I also want to send an ACK back to App1 as soon as App2
receives M1, i.e. I want App2 to send two types of ACKs to App1:
1. An ACK that says that App2 received M1
2. An ACK after M1 is processed.

Can I make App1 use Dealer socket and App2 Router ? How will the matching
happen on App1's end ? If this pattern is not right, could you please point
out the right pattern to use for this use case.


I greatly appreciate your replies.

Many thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/cd463ffd/attachment.htm 

From andrew at research.att.com  Thu Jul 12 20:51:14 2012
From: andrew at research.att.com (Andrew Hume)
Date: Thu, 12 Jul 2012 11:51:14 -0700
Subject: [zeromq-dev] Quick question on usage of patterns
In-Reply-To: <CAKni0ed721gwACAzS16W_fakGLx1A2oSBDZAJdXc5H--MfrXiA@mail.gmail.com>
References: <CAKni0ed721gwACAzS16W_fakGLx1A2oSBDZAJdXc5H--MfrXiA@mail.gmail.com>
Message-ID: <35383EE6-5A62-4008-BF9E-E48163CD7E61@research.att.com>

unless you're not telling us something,
the simplest solution is two sockets:
app1.PUSH => app2.PULL: for M1
app2.PUSH => app1.PULL: for various ACKs

as for the matching, i would include an job id with M1 and the ACKs.
or if you don't want to do that, use the system name (Box2) as the id.
both these solutions generalise in different directions.

	andrew

On Jul 12, 2012, at 11:23 AM, Diffuser78 wrote:

> Hello All,
> 
> I need some suggestion on my design and ways in which it can be done using ZMQ sockets.
> 
> 
> I have an app  App1 on Box1. It sends a message M1 (128K) to an app App2 on Box2. App2 will do some processing on M1 before sending an ACK back. In my application, I also want to send an ACK back to App1 as soon as App2 receives M1, i.e. I want App2 to send two types of ACKs to App1:
> 1. An ACK that says that App2 received M1
> 2. An ACK after M1 is processed.
> 
> Can I make App1 use Dealer socket and App2 Router ? How will the matching happen on App1's end ? If this pattern is not right, could you please point out the right pattern to use for this use case.
> 
> 
> I greatly appreciate your replies.
> 
> Many thanks.
> 
>  
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


------------------
Andrew Hume  (best -> Telework) +1 623-551-2845
andrew at research.att.com  (Work) +1 973-236-2014
AT&T Labs - Research; member of USENIX and LOPSA




-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/72c7895c/attachment.htm 

From lists at chuckremes.com  Thu Jul 12 20:56:02 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Thu, 12 Jul 2012 13:56:02 -0500
Subject: [zeromq-dev] Quick question on usage of patterns
In-Reply-To: <CAKni0ed721gwACAzS16W_fakGLx1A2oSBDZAJdXc5H--MfrXiA@mail.gmail.com>
References: <CAKni0ed721gwACAzS16W_fakGLx1A2oSBDZAJdXc5H--MfrXiA@mail.gmail.com>
Message-ID: <93506029-8A21-4DEA-8AD3-6B02A7AAB039@chuckremes.com>

On Jul 12, 2012, at 1:23 PM, Diffuser78 wrote:

> Hello All,
> 
> I need some suggestion on my design and ways in which it can be done using ZMQ sockets.
> 
> 
> I have an app  App1 on Box1. It sends a message M1 (128K) to an app App2 on Box2. App2 will do some processing on M1 before sending an ACK back. In my application, I also want to send an ACK back to App1 as soon as App2 receives M1, i.e. I want App2 to send two types of ACKs to App1:
> 1. An ACK that says that App2 received M1
> 2. An ACK after M1 is processed.
> 
> Can I make App1 use Dealer socket and App2 Router ? How will the matching happen on App1's end ? If this pattern is not right, could you please point out the right pattern to use for this use case.

If you want to generalize this so that App1 can talk to App2, App3, App4 and App5 and each one can respond back properly, I would suggest using Dealer (on App1) and Router sockets on the remaining Apps. See the write up on the wiki (http://www.zeromq.org/tutorials:xreq-and-xrep) regarding how to format messages for talking between Dealer/Router (renamed from XREQ/XREP).

cr


From jhawk28 at gmail.com  Fri Jul 13 00:53:57 2012
From: jhawk28 at gmail.com (Joshua Foster)
Date: Thu, 12 Jul 2012 18:53:57 -0400
Subject: [zeromq-dev] Broadcasting with a Router ?
In-Reply-To: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
References: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
Message-ID: <4FFF5585.2090301@gmail.com>

You technically could if you kept track of all the addresses that you 
received. The main downside is that you don't know when someone 
connects/disconnects. You only know what requests were received.

Joshua

> Gr?goire Passault <mailto:g.passault at gmail.com>
> Thursday, July 12, 2012 9:43 AM
> Hi,
>
> Since I'm new to ZeroMQ I really appologize if my question is stupid, 
> but can we use a Router to broadcast messages to all clients ?
>
> I mean, I don't see any technical reason it couldn't, and I know I 
> also could create Request/Req and Pub/Sub sockets for instance, but I 
> don't want to create two of them if one can do the job.
>
> Best regards,
> Gr?goire Passaukt
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/3d7c5732/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: postbox-contact.jpg
Type: image/jpeg
Size: 1032 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/3d7c5732/attachment.jpg 

From g.passault at gmail.com  Fri Jul 13 00:57:06 2012
From: g.passault at gmail.com (=?ISO-8859-1?Q?Gr=E9goire_Passault?=)
Date: Fri, 13 Jul 2012 00:57:06 +0200
Subject: [zeromq-dev] Broadcasting with a Router ?
In-Reply-To: <CADF6OR6dO6wgsYVutoaoHvh4+hSKCc=5SXhiqUJj5DAw6qKZmg@mail.gmail.com>
References: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
	<4FFF5585.2090301@gmail.com>
	<CADF6OR6dO6wgsYVutoaoHvh4+hSKCc=5SXhiqUJj5DAw6qKZmg@mail.gmail.com>
Message-ID: <CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>

Of course ...

But think of an "hybrid" Router-and-Pub socket, users could subscribe to
broadcast or exchange adressed data in the same socket. Why not?

Greg
Le 13 juil. 2012 00:54, "Joshua Foster" <jhawk28 at gmail.com> a ?crit :
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120713/613555c0/attachment.htm 

From jhawk28 at gmail.com  Fri Jul 13 01:17:36 2012
From: jhawk28 at gmail.com (Joshua Foster)
Date: Thu, 12 Jul 2012 19:17:36 -0400
Subject: [zeromq-dev] Broadcasting with a Router ?
In-Reply-To: <CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>
References: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
	<4FFF5585.2090301@gmail.com>
	<CADF6OR6dO6wgsYVutoaoHvh4+hSKCc=5SXhiqUJj5DAw6qKZmg@mail.gmail.com>
	<CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>
Message-ID: <4FFF5B10.2020307@gmail.com>

Anything is possible :)

There was some experimentation going on the 3.0 version that created a 
Generic socket. Your hybrid socket could be developed with it because 
you would get the Connect/Disconnect information.

Joshua

> Gr?goire Passault <mailto:g.passault at gmail.com>
> Thursday, July 12, 2012 6:57 PM
>
> Of course ...
>
> But think of an "hybrid" Router-and-Pub socket, users could subscribe 
> to broadcast or exchange adressed data in the same socket. Why not?
>
> Greg
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> Gr?goire Passault <mailto:g.passault at gmail.com>
> Thursday, July 12, 2012 9:43 AM
> Hi,
>
> Since I'm new to ZeroMQ I really appologize if my question is stupid, 
> but can we use a Router to broadcast messages to all clients ?
>
> I mean, I don't see any technical reason it couldn't, and I know I 
> also could create Request/Req and Pub/Sub sockets for instance, but I 
> don't want to create two of them if one can do the job.
>
> Best regards,
> Gr?goire Passaukt
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/ebe3601b/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: postbox-contact.jpg
Type: image/jpeg
Size: 1032 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/ebe3601b/attachment.jpg 

From thatch45 at gmail.com  Fri Jul 13 01:37:41 2012
From: thatch45 at gmail.com (Thomas S Hatch)
Date: Thu, 12 Jul 2012 17:37:41 -0600
Subject: [zeromq-dev] mailbox.cpp Assertion
In-Reply-To: <CAEYnEwFJ=KO30p_HpMWyN8USAzJQ38tt=j1E86DdXHLszub=JQ@mail.gmail.com>
References: <CAO18cg9Z4-tvhNNGqEWjg2sH2cQAJDtm5UrJDwCacZU3PkzyUw@mail.gmail.com>
	<CADL5_sj9ayjrU3zNOyQ=LCObN9Hg2z_WP1oOZ=y=EgQ3OismWw@mail.gmail.com>
	<CAO18cg-853ZzJJqGrm4ON5nxUi6-C-LN0QWfP4pVS1+q1ThDzg@mail.gmail.com>
	<CAEYnEwEq-JCmJksTCu90HyfWN5m4zZb_uNmHOM5ZdB81AgmuyA@mail.gmail.com>
	<CAO18cg8VeDNt77vZ_MJUa_TU478kYw6cDgzFtxenwb9XcL5Ggg@mail.gmail.com>
	<D0F161A9-9527-40C7-9FC2-396C09CF71D5@chuckremes.com>
	<CAO18cg_jZSAkn+-em0r7zk0VtoO98uW7aSD7MG0muPviZJKEag@mail.gmail.com>
	<B9A75763-5B12-40A6-86C5-8F630C4BC681@chuckremes.com>
	<CAO18cg_=FK_FSjWXF4o7rK5a_UiKGuLCGpBeree6UXBapk+ygg@mail.gmail.com>
	<CAO18cg-s5WFaWQLFbYSTGeg2F-vPtApR8cvc9BhLCfGiMARhww@mail.gmail.com>
	<CAEYnEwFJ=KO30p_HpMWyN8USAzJQ38tt=j1E86DdXHLszub=JQ@mail.gmail.com>
Message-ID: <CAO18cg_4-q-q2Ab=0+QAk2XsC+_YDXf_MopQ-79UhX8kN-PgUQ@mail.gmail.com>

Thanks guys,

Unfortunately there is no python stack trace, it looks like the error is
coming from a zeromq thread and then sending the SIGABRT up. I don't know
where the conflict is coming from though. I guess I need a better
understanding of some of the threading that zeromq is doing underneath the
python

On Mon, Jul 9, 2012 at 2:34 PM, Ian Barber <ian.barber at gmail.com> wrote:

>
>
> On Mon, Jul 9, 2012 at 9:26 PM, Thomas S Hatch <thatch45 at gmail.com> wrote:
>>
>>  Actually, I should ask...
>>
>> The main issue we are seeing in here is that the SUB socket looses
>> connection from the PUB socket when connecting over unreliable connections
>> like the internet. Is there a way to detect if the SUB socket is still
>> connected to the PUB socket?
>>
>
> Sort of: Either layer a heartbeat on top, or use the ZMQ_MONITOR stuff in
> ZMQ3.1+ to look for disconnect events. I suspect you'll get better accuracy
> with the heartbeat (as it'll catch a multitude of sins), but monitor would
> easier.
>
> Ian
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120712/5bfaa226/attachment.htm 

From ian.barber at gmail.com  Fri Jul 13 03:14:50 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Fri, 13 Jul 2012 02:14:50 +0100
Subject: [zeromq-dev] Broadcasting with a Router ?
In-Reply-To: <CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>
References: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
	<4FFF5585.2090301@gmail.com>
	<CADF6OR6dO6wgsYVutoaoHvh4+hSKCc=5SXhiqUJj5DAw6qKZmg@mail.gmail.com>
	<CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>
Message-ID: <CAEYnEwG7TctZ3WL_8tnDNxDW_gXHGoap1UMdH8VOrWGOuQnDAw@mail.gmail.com>

On Thu, Jul 12, 2012 at 11:57 PM, Gr?goire Passault <g.passault at gmail.com>wrote:

> Of course ...
>
> But think of an "hybrid" Router-and-Pub socket, users could subscribe to
> broadcast or exchange adressed data in the same socket. Why not?
>
> Greg
>

You could build it with PUB pretty easily, have every subscriber subscribe
to some string for their identity, and to an 'all' flag, then have the
publisher send to whichever it preferred.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120713/7b8592f1/attachment.htm 

From g.passault at gmail.com  Fri Jul 13 12:05:01 2012
From: g.passault at gmail.com (=?ISO-8859-1?Q?Gr=E9goire_Passault?=)
Date: Fri, 13 Jul 2012 12:05:01 +0200
Subject: [zeromq-dev] Broadcasting with a Router ?
In-Reply-To: <CAEYnEwG7TctZ3WL_8tnDNxDW_gXHGoap1UMdH8VOrWGOuQnDAw@mail.gmail.com>
References: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
	<4FFF5585.2090301@gmail.com>
	<CADF6OR6dO6wgsYVutoaoHvh4+hSKCc=5SXhiqUJj5DAw6qKZmg@mail.gmail.com>
	<CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>
	<CAEYnEwG7TctZ3WL_8tnDNxDW_gXHGoap1UMdH8VOrWGOuQnDAw@mail.gmail.com>
Message-ID: <CADF6OR4fSQf8uwSzaF-82ci0hm3Ry=nTjsOfqJMjkiio=4uRZQ@mail.gmail.com>

Yes but a subscriber can't send any message, right?

Greg
Le 13 juil. 2012 03:15, "Ian Barber" <ian.barber at gmail.com> a ?crit :

>
>
> On Thu, Jul 12, 2012 at 11:57 PM, Gr?goire Passault <g.passault at gmail.com>wrote:
>
>> Of course ...
>>
>> But think of an "hybrid" Router-and-Pub socket, users could subscribe to
>> broadcast or exchange adressed data in the same socket. Why not?
>>
>> Greg
>>
>
> You could build it with PUB pretty easily, have every subscriber subscribe
> to some string for their identity, and to an 'all' flag, then have the
> publisher send to whichever it preferred.
>
> Ian
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120713/83d1847a/attachment.htm 

From mortenmoellerriis at gmail.com  Fri Jul 13 14:04:41 2012
From: mortenmoellerriis at gmail.com (=?iso-8859-1?Q?Morten_M=F8ller_Riis?=)
Date: Fri, 13 Jul 2012 14:04:41 +0200
Subject: [zeromq-dev] ruby: rbzmq.c:99: context_free: Assertion `rc == 0'
	failed.
Message-ID: <D39B56D8-8CF3-4C76-A246-558F7DE04720@gmail.com>

Hi!

I am seeing this error sometimes with a daemon that I created which uses ZMQ.

ruby: rbzmq.c:99: context_free: Assertion `rc == 0' failed.

I am using the 'zmq' gem on Ruby 1.9.2p0 with zeromq-2.1.11.


The program is working perfectly until this error seems to occur. It happens after the following error from a STDOUT#flush call. But I am not sure whether they would be related.

/usr/local/minion/lib/kernel_ext.rb:24:in `flush': deadlock; recursive locking (ThreadError)

Any pointers would be highly appreciated!

Best regards
Morten M?ller Riis




From Zafar.Iqbal at lnties.com  Fri Jul 13 15:00:42 2012
From: Zafar.Iqbal at lnties.com (Zafar Iqbal)
Date: Fri, 13 Jul 2012 13:00:42 +0000
Subject: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX
Message-ID: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>

Hi All,

I am trying to compile Zeromq-2.1.9 in QNX self hosted system, I was getting the following errors.

1. Initially after running the configuration file, Makefile generated sucessfully. When I invoked make , the compilation started and when it reached to the linking stage it has given the error -lcrypto not found, just to chek it I removed -lcrypto from the makefile then it has shown the error -lsocket not found, When I looked into the libs path I found both the were available.

After some analysis I found -rpath was not set, I set it to /usr/lib, after that programme compiled successfully.


But after that I faced the Following problem:

2. When I tried to install it giving make install:
I got the following error.


Making install in src
make[1]: Entering directory `/root/zeromq-2.1.9/src'
make[2]: Entering directory `/root/zeromq-2.1.9/src'
test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
 /bin/sh ../libtool   --mode=install ../config/install-sh -c   libzmq.la '/usr/loca
libtool: install: ../config/install-sh -c .libs/libzmq.lai /usr/local/lib/libzmq.la
../config/install-sh: .libs/libzmq.lai does not exist.
make[2]: *** [install-libLTLIBRARIES] Error 1
make[2]: Leaving directory `/root/zeromq-2.1.9/src'
make[1]: *** [install-am] Error 2
make[1]: Leaving directory `/root/zeromq-2.1.9/src'
make: *** [install-recursive] Error 1


I am not able to find out what is the actual problem, How to solve the same.

Please let me know if any of you have gone through the similar kind of problem.



Thanking a lot.

Larsen & Toubro Limited

www.larsentoubro.com

This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.

From armin at steinhoff.de  Fri Jul 13 15:28:19 2012
From: armin at steinhoff.de (Armin Steinhoff)
Date: Fri, 13 Jul 2012 15:28:19 +0200
Subject: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX
In-Reply-To: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>
References: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>
Message-ID: <50002273.2060402@steinhoff.de>

Zafar Iqbal wrote:
> Hi All,
>
> I am trying to compile Zeromq-2.1.9 in QNX self hosted system, I was getting the following errors.
>
> 1. Initially after running the configuration file, Makefile generated sucessfully. When I invoked make , the compilation started and when it reached to the linking stage it has given the error -lcrypto not found, just to chek it I removed -lcrypto from the makefile then it has shown the error -lsocket not found, When I looked into the libs path I found both the were available.

  .... and what is your LD_LIBRARY_PATH ? Does it include /usr/lib ?

>
> After some analysis I found -rpath was not set, I set it to /usr/lib, after that programme compiled successfully.

  ... not sure if the -rpath option works with QCC front end! Are you 
using QNX65 ?

--Armin

http://steinhoff-automation.com

>
>
> But after that I faced the Following problem:
>
> 2. When I tried to install it giving make install:
> I got the following error.
>
>
> Making install in src
> make[1]: Entering directory `/root/zeromq-2.1.9/src'
> make[2]: Entering directory `/root/zeromq-2.1.9/src'
> test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
>   /bin/sh ../libtool   --mode=install ../config/install-sh -c   libzmq.la '/usr/loca
> libtool: install: ../config/install-sh -c .libs/libzmq.lai /usr/local/lib/libzmq.la
> ../config/install-sh: .libs/libzmq.lai does not exist.
> make[2]: *** [install-libLTLIBRARIES] Error 1
> make[2]: Leaving directory `/root/zeromq-2.1.9/src'
> make[1]: *** [install-am] Error 2
> make[1]: Leaving directory `/root/zeromq-2.1.9/src'
> make: *** [install-recursive] Error 1
>
>
> I am not able to find out what is the actual problem, How to solve the same.
>
> Please let me know if any of you have gone through the similar kind of problem.
>
>
>
> Thanking a lot.
>
> Larsen & Toubro Limited
>
> www.larsentoubro.com
>
> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



From Zafar.Iqbal at lnties.com  Fri Jul 13 19:16:04 2012
From: Zafar.Iqbal at lnties.com (Zafar Iqbal)
Date: Fri, 13 Jul 2012 17:16:04 +0000
Subject: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX
In-Reply-To: <50002273.2060402@steinhoff.de>
References: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>,
	<50002273.2060402@steinhoff.de>
Message-ID: <7E56B0710035754D80A22F7BA63062C806DA777B@POCITMSEXMB01.LntUniverse.com>



> Hi All,
>
> I am trying to compile Zeromq-2.1.9 in QNX self hosted system, I was getting the following errors.
>
> 1. Initially after running the configuration file, Makefile generated sucessfully. When I invoked make , the compilation started and when it reached to the linking stage it has given the error -lcrypto not found, just to chek it I removed -lcrypto from the makefile then it has shown the error -lsocket not found, When I looked into the libs path I found both the were available.

  .... and what is your LD_LIBRARY_PATH ? Does it include /usr/lib ?

Yes my LD_LIBRARY_PATH includes /usr/lib

>
> After some analysis I found -rpath was not set, I set it to /usr/lib, after that programme compiled successfully.

  ... not sure if the -rpath option works with QCC front end! Are you
using QNX65 ?

Yes I am using QNX65.



http://steinhoff-automation.com

>
>
> But after that I faced the Following problem:
>
> 2. When I tried to install it giving make install:
> I got the following error.
>
>
> Making install in src
> make[1]: Entering directory `/root/zeromq-2.1.9/src'
> make[2]: Entering directory `/root/zeromq-2.1.9/src'
> test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
>   /bin/sh ../libtool   --mode=install ../config/install-sh -c   libzmq.la '/usr/loca
> libtool: install: ../config/install-sh -c .libs/libzmq.lai /usr/local/lib/libzmq.la
> ../config/install-sh: .libs/libzmq.lai does not exist.
> make[2]: *** [install-libLTLIBRARIES] Error 1
> make[2]: Leaving directory `/root/zeromq-2.1.9/src'
> make[1]: *** [install-am] Error 2
> make[1]: Leaving directory `/root/zeromq-2.1.9/src'
> make: *** [install-recursive] Error 1
>
>
> I am not able to find out what is the actual problem, How to solve the same.
>
> Please let me know if any of you have gone through the similar kind of problem.
>
>
>
> Thanking a lot.
>
> Larsen & Toubro Limited
>
> www.larsentoubro.com
>
> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>


Larsen & Toubro Limited

www.larsentoubro.com

This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.

From armin at steinhoff.de  Fri Jul 13 20:36:25 2012
From: armin at steinhoff.de (Armin Steinhoff)
Date: Fri, 13 Jul 2012 20:36:25 +0200
Subject: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX
In-Reply-To: <7E56B0710035754D80A22F7BA63062C806DA777B@POCITMSEXMB01.LntUniverse.com>
References: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>,
	<50002273.2060402@steinhoff.de>
	<7E56B0710035754D80A22F7BA63062C806DA777B@POCITMSEXMB01.LntUniverse.com>
Message-ID: <50006AA9.9010008@steinhoff.de>


Hi,

you should use "bash" as shell. Download the sources of bash ... it can 
be compiled without problem.
The ksh or sh makes often problems with bash based scripts/commands.

--Armin

Zafar Iqbal wrote:
>
>> Hi All,
>>
>> I am trying to compile Zeromq-2.1.9 in QNX self hosted system, I was getting the following errors.
>>
>> 1. Initially after running the configuration file, Makefile generated sucessfully. When I invoked make , the compilation started and when it reached to the linking stage it has given the error -lcrypto not found, just to chek it I removed -lcrypto from the makefile then it has shown the error -lsocket not found, When I looked into the libs path I found both the were available.
>    .... and what is your LD_LIBRARY_PATH ? Does it include /usr/lib ?
>
> Yes my LD_LIBRARY_PATH includes /usr/lib
>
>> After some analysis I found -rpath was not set, I set it to /usr/lib, after that programme compiled successfully.
>    ... not sure if the -rpath option works with QCC front end! Are you
> using QNX65 ?
>
> Yes I am using QNX65.
>
>
>
> http://steinhoff-automation.com
>
>>
>> But after that I faced the Following problem:
>>
>> 2. When I tried to install it giving make install:
>> I got the following error.
>>
>>
>> Making install in src
>> make[1]: Entering directory `/root/zeromq-2.1.9/src'
>> make[2]: Entering directory `/root/zeromq-2.1.9/src'
>> test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
>>    /bin/sh ../libtool   --mode=install ../config/install-sh -c   libzmq.la '/usr/loca
>> libtool: install: ../config/install-sh -c .libs/libzmq.lai /usr/local/lib/libzmq.la
>> ../config/install-sh: .libs/libzmq.lai does not exist.
>> make[2]: *** [install-libLTLIBRARIES] Error 1
>> make[2]: Leaving directory `/root/zeromq-2.1.9/src'
>> make[1]: *** [install-am] Error 2
>> make[1]: Leaving directory `/root/zeromq-2.1.9/src'
>> make: *** [install-recursive] Error 1
>>
>>
>> I am not able to find out what is the actual problem, How to solve the same.
>>
>> Please let me know if any of you have gone through the similar kind of problem.
>>
>>
>>
>> Thanking a lot.
>>
>> Larsen & Toubro Limited
>>
>> www.larsentoubro.com
>>
>> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
> Larsen & Toubro Limited
>
> www.larsentoubro.com
>
> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
>



From armin at steinhoff.de  Fri Jul 13 22:04:25 2012
From: armin at steinhoff.de (Armin Steinhoff)
Date: Fri, 13 Jul 2012 22:04:25 +0200
Subject: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX
In-Reply-To: <7E56B0710035754D80A22F7BA63062C806DA777B@POCITMSEXMB01.LntUniverse.com>
References: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>,
	<50002273.2060402@steinhoff.de>
	<7E56B0710035754D80A22F7BA63062C806DA777B@POCITMSEXMB01.LntUniverse.com>
Message-ID: <50007F49.8090502@steinhoff.de>


Hi,

I have recompiled zmq 2.2 and configured with

CC=qcc CXX=QCC ./configure
make
make install

without any problem for QNX 6.5 SP1 by using "bash".

BR

--Armin



Zafar Iqbal wrote:
>
>> Hi All,
>>
>> I am trying to compile Zeromq-2.1.9 in QNX self hosted system, I was getting the following errors.
>>
>> 1. Initially after running the configuration file, Makefile generated sucessfully. When I invoked make , the compilation started and when it reached to the linking stage it has given the error -lcrypto not found, just to chek it I removed -lcrypto from the makefile then it has shown the error -lsocket not found, When I looked into the libs path I found both the were available.
>    .... and what is your LD_LIBRARY_PATH ? Does it include /usr/lib ?
>
> Yes my LD_LIBRARY_PATH includes /usr/lib
>
>> After some analysis I found -rpath was not set, I set it to /usr/lib, after that programme compiled successfully.
>    ... not sure if the -rpath option works with QCC front end! Are you
> using QNX65 ?
>
> Yes I am using QNX65.
>
>
>
> http://steinhoff-automation.com
>
>>
>> But after that I faced the Following problem:
>>
>> 2. When I tried to install it giving make install:
>> I got the following error.
>>
>>
>> Making install in src
>> make[1]: Entering directory `/root/zeromq-2.1.9/src'
>> make[2]: Entering directory `/root/zeromq-2.1.9/src'
>> test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
>>    /bin/sh ../libtool   --mode=install ../config/install-sh -c   libzmq.la '/usr/loca
>> libtool: install: ../config/install-sh -c .libs/libzmq.lai /usr/local/lib/libzmq.la
>> ../config/install-sh: .libs/libzmq.lai does not exist.
>> make[2]: *** [install-libLTLIBRARIES] Error 1
>> make[2]: Leaving directory `/root/zeromq-2.1.9/src'
>> make[1]: *** [install-am] Error 2
>> make[1]: Leaving directory `/root/zeromq-2.1.9/src'
>> make: *** [install-recursive] Error 1
>>
>>
>> I am not able to find out what is the actual problem, How to solve the same.
>>
>> Please let me know if any of you have gone through the similar kind of problem.
>>
>>
>>
>> Thanking a lot.
>>
>> Larsen & Toubro Limited
>>
>> www.larsentoubro.com
>>
>> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
> Larsen & Toubro Limited
>
> www.larsentoubro.com
>
> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
>



From ian.barber at gmail.com  Fri Jul 13 22:24:35 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Fri, 13 Jul 2012 21:24:35 +0100
Subject: [zeromq-dev] Broadcasting with a Router ?
In-Reply-To: <CADF6OR4fSQf8uwSzaF-82ci0hm3Ry=nTjsOfqJMjkiio=4uRZQ@mail.gmail.com>
References: <CADF6OR6f+X1VYXd_YuhSCJQ-8Qqnb6jFKaisfvEoH8OeGOEwDg@mail.gmail.com>
	<4FFF5585.2090301@gmail.com>
	<CADF6OR6dO6wgsYVutoaoHvh4+hSKCc=5SXhiqUJj5DAw6qKZmg@mail.gmail.com>
	<CADF6OR5z2oU032sSU17G-a6vwMOyUdNs14DBxMaF-3SCL3hufg@mail.gmail.com>
	<CAEYnEwG7TctZ3WL_8tnDNxDW_gXHGoap1UMdH8VOrWGOuQnDAw@mail.gmail.com>
	<CADF6OR4fSQf8uwSzaF-82ci0hm3Ry=nTjsOfqJMjkiio=4uRZQ@mail.gmail.com>
Message-ID: <CAEYnEwE8-fBLeG3e-rB5oQytAQwxM=HEBJJvAaJ7MijPbb9FYg@mail.gmail.com>

On Fri, Jul 13, 2012 at 11:05 AM, Gr?goire Passault <g.passault at gmail.com>wrote:

> Yes but a subscriber can't send any message, right?
>
> Greg
>

Well, technically they can send subscribe messages on 3.1+, but that is
probably pushing it too far :)

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120713/ae79dbd0/attachment.htm 

From rtummala at appdynamics.com  Sat Jul 14 00:35:43 2012
From: rtummala at appdynamics.com (Rama Tummala)
Date: Fri, 13 Jul 2012 15:35:43 -0700
Subject: [zeromq-dev] Question on client side message queueing
Message-ID: <CAOdTiWcGFwnrUef97h0VxKrC78uGrw6RhLiZGMxAkHrsNQzM+A@mail.gmail.com>

I have a N-1 fan in model. I am using ROUTER-DEALER-REP.  (I use ZMQ.Poll
to pull the requests from the ROUTER socket and push to DEALER socket.)
The ROUTER and DEALER sockets bind at the server and client simply connects
to the ROUTER to push requests.
In my case, around 200 clients connect to the ROUTER and send multiple
requests.
I am trying to make best use of the client side message queue feature
provided by zeromq.

In case of a REQ-REP its very clear. The message stays in the message queue
(on the client end) until the REP socket is ready to process.
How does this concept translate to a more complicated model like
REQ-ROUTER-DEALER-REP.
In the doc, it also says that the queueing happens closer to the receiver.

I am trying to avoid the situation where all the messages are queued up on
the server. (For example, 200 clients * 100 messages queued up).
Any help is much appreciated.



Regards
Rama
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120713/e434f8fa/attachment.htm 

From ph at imatix.com  Sat Jul 14 11:01:58 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Sat, 14 Jul 2012 18:01:58 +0900
Subject: [zeromq-dev] [czmq] src/zsocket.c,
 this looks like unused variable to me
In-Reply-To: <1342068143.17421.6.camel@laure.segg2.videotron.ca>
References: <1342068143.17421.6.camel@laure.segg2.videotron.ca>
Message-ID: <CADL5_shprCWtZc86ZVBxOOebLGT454n+XJ-SV43j_AkG-qGMig@mail.gmail.com>

On Thu, Jul 12, 2012 at 1:42 PM, Gilles J. Seguin <segg2 at videotron.ca> wrote:

> fedora gcc-4.7.0-5.fc17.x86_64
> $ gcc -I../include -Werror -Wno-unused -Wall -c zsocket.c -o zsocket.o
> zsocket.c: In function ?zsocket_poll?:
> zsocket.c:152:9: error: unused variable ?rc? [-Werror=unused-variable]
> cc1: all warnings being treated as errors

Indeed, it's unused. Some other random compiler was complaining about
"return value discarded".

Feel free to send a pull request to fix this.

-Pieter

From aalok at ideadevice.com  Sat Jul 14 20:52:03 2012
From: aalok at ideadevice.com (aalok sood)
Date: Sun, 15 Jul 2012 00:22:03 +0530
Subject: [zeromq-dev] A query about implementing aborts in request reply
	broker
Message-ID: <CAG+SKy5t68xpGrgFdFoWoz9SjDZv9XjuK8A0-TX8OYTGStbdtg@mail.gmail.com>

I have a request reply broker and it does the job really well.

Now if I want to tell my workers to abort a bunch of these requests, based
on some request parameter.

For this I can use a pub-sub, but how do I collect the results?
Or should I try to keep a mapping in my broker about which worker is
running which request.
Is this some awesomeness of zeromq that I am blind to?

--
Aalok
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120715/664de341/attachment.htm 

From vickyqasim at gmail.com  Sun Jul 15 04:13:38 2012
From: vickyqasim at gmail.com (Vicky Qasim)
Date: Sun, 15 Jul 2012 07:43:38 +0530
Subject: [zeromq-dev] Problem About Installing BASH in QNX
Message-ID: <CAN3YSa_9gmc0Zg-izFOhjJOcgoJf_4428dNVc_Z4xBAEdx77Xg@mail.gmail.com>

HI,

I am using QNX 6.5.0, I am trying to Install BASH into it I have downloaded
bash 4.0 and trying to install into the qnx system.

I ran env CC=qcc ./configure

after that make

But I am getting the following error.

fpurge.c:119:3: error: #error "Please port gnulib fpurge.c to your
platform! Look at the
definitions of fflush, setvbuf and ungetc on your system, then report
this to bug-gnulib.
cc: /usr/qnx650/host/qnx6/x86/usr/
lib/gcc/i486-pc-nto-qnx6.5.0/4.4.2/cc1 error 1
make[1]: *** [fpurge.o] Error 1
make[1]: Leaving directory `/root/bash-4.0-rc1/lib/sh'
make: *** [lib/sh/libsh.a] Error 1

Please let me know if any one is having the solution of the same.

Thanks in Advance
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120715/29f3b68f/attachment.htm 

From Sharon.Ben-Asher at avg.com  Sun Jul 15 14:38:09 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Sun, 15 Jul 2012 12:38:09 +0000
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the Android APK
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D505B@ex10czdmb01>

Hello,

I managed to compile the ZeroMQ + Java binding shared objects for ARM.
Thias is what file gives me on the two so files
libzmq.so.3.0.0: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), dynamically linked, not stripped
libjzmq.so: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), dynamically linked, not stripped
and of course, I also have zmq.jar
now I created a new Android project in Eclipse.
Following advice on the net:

*        I placed the two so files udner libs/armeabi (renamed libzmq.so.3.0.0 into libzmq.so)

*        I placed the jar file under libs, and added to the build path

*        Took the tripping sample and called it from main activity

*        Compile was successful. However, debugging the apk on a mobile device, I get the following error during System.loadLibrary in the ctor of ZMQ
file name: /data/data/com.avgmobilation.zeromq/lib/libjzmq.so
Cannot load library: reloc_library[1289]:   157 cannot locate '_ZNSs6assignEPKcj'...
and later on this gets translated into an UnsatisfiedLinkError

Even tried to build the apk on the linux machine where the so was compiled - same error.
Any help is appreciated...

Thanks,

Sharon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120715/e8ea844d/attachment.htm 

From ian.barber at gmail.com  Sun Jul 15 17:40:18 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Sun, 15 Jul 2012 16:40:18 +0100
Subject: [zeromq-dev] A query about implementing aborts in request reply
	broker
In-Reply-To: <CAG+SKy5t68xpGrgFdFoWoz9SjDZv9XjuK8A0-TX8OYTGStbdtg@mail.gmail.com>
References: <CAG+SKy5t68xpGrgFdFoWoz9SjDZv9XjuK8A0-TX8OYTGStbdtg@mail.gmail.com>
Message-ID: <CAEYnEwE6-VwvMyikB+PFeTB9kS_mNf5qLgubKL0iy6eq0Q+05Q@mail.gmail.com>

On Sat, Jul 14, 2012 at 7:52 PM, aalok sood <aalok at ideadevice.com> wrote:

> I have a request reply broker and it does the job really well.
>
> Now if I want to tell my workers to abort a bunch of these requests, based
> on some request parameter.
>
> For this I can use a pub-sub, but how do I collect the results?
>

I would use a pub-sub and a pull or similar socket - so requests to cancel
are published, and you get a confirm message when they've been cancelled
sent back over the pull socket. That said, given you req-rep at the moment,
I'm guessing your work distributor is a dealer or router socket, so you
could collect the results directly through that.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120715/9c3c7b61/attachment.htm 

From stefan at konink.de  Sun Jul 15 20:00:47 2012
From: stefan at konink.de (Stefan de Konink)
Date: Sun, 15 Jul 2012 20:00:47 +0200
Subject: [zeromq-dev] High memory usage (memory leak?)
Message-ID: <5003054F.8040202@konink.de>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Hi,

We are using ZeroMQ 2.2 and are quite happy about it. We wish we could
move to ZeroMQ 3 but some outstanding bugs regarding the PubSub
implementation prevent this.

One of our current mysteries is the enormous high amount of memory our
pubsub uses. The clients on the other hand consume a very high virtual
memory amount. (Think in order of 500MB+)

It seems that once in a few weeks we end up with a:

FATAL ERROR: OUT OF MEMORY (yqueue.hpp:54)


Our source code is pretty simple - no mallocs from our side what so
ever and every init is closed; Still if I look to the virtual memory
and real memory of one of my clients I see:

146MB virtual
56256 real

...where is this memory spend?


Stefan




#include <stdio.h>
#include <stdlib.h>
#include <assert.h>
#include <zmq.h>

int main (int argc, char *argv[]) {
    if (argc != 3) {
        printf("%s [receiver] [pubsub]\n\nEx.\n%s tcp://127.0.0.1:7807
tcp://127.0.0.1:7817\n", argv[0], argv[0]);
        exit(-1);
    }

        void *context  = zmq_init (1);
        void *pubsub   = zmq_socket (context, ZMQ_PUB);
        void *receiver = zmq_socket (context, ZMQ_PULL);

        zmq_bind (pubsub,   argv[2]);
        zmq_bind (receiver, argv[1]);

        while (1) {
                int64_t more;
                size_t more_size = sizeof more;
                do {
                        /* Create an empty 0MQ message to hold the
message part */
                        zmq_msg_t part;
                        int rc = zmq_msg_init (&part);
                        assert (rc == 0);
                        /* Block until a message is available to be
received from socket */
                        rc = zmq_recv (receiver, &part, 0);
                        assert (rc == 0);
                        /* Determine if more message parts are to
follow */
                        rc = zmq_getsockopt (receiver, ZMQ_RCVMORE,
&more, &more_size);
                        assert (rc == 0);
                        if (more) {
                                zmq_send (pubsub, &part, ZMQ_SNDMORE);
                        } else {
                                zmq_send (pubsub, &part, 0);
                        }
                        zmq_msg_close (&part);
                } while (more);
        }
}
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.18 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEAREKAAYFAlADBU4ACgkQYH1+F2Rqwn3iVwCeLZGRgTRVVHXVAL9smKLYnEbC
+LkAoIcWRhK1YylyLUlpkDEtoLgF0qiS
=Hy+T
-----END PGP SIGNATURE-----

From hurtonm at gmail.com  Sun Jul 15 20:03:55 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Sun, 15 Jul 2012 20:03:55 +0200
Subject: [zeromq-dev] Websockets as a Transport ?
In-Reply-To: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
References: <024001cd5a13$16fa8e10$44efaa30$@petrolink.com>
Message-ID: <CALdZE=u5-GZ-ahqCZcp8QeAWxyPUk4QQvvqujd7pwiZpTwj4+w@mail.gmail.com>

Hi Mark, I am working on adding websocket transport into 0MQ.
Will post a message once its ready for testing.

- Martin

On Wed, Jul 4, 2012 at 8:30 PM, Mark Farnan <mark.farnan at petrolink.com> wrote:
> Howdy all,
>
>
>
> I?m evaluating zeroMQ, and one main criteria we have is to support
> Websockets for client to server communications.
>
>
>
> Has there been any work on making a native Websocket transport for ZeroMQ in
> the core C++ libraries ?
>
>
>
>
>
> Regards
>
>
>
> Mark Farnan
>
> Director, Product Development
>
> Petrolink Services Inc
>
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From hurtonm at gmail.com  Mon Jul 16 00:10:45 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Mon, 16 Jul 2012 00:10:45 +0200
Subject: [zeromq-dev] High memory usage (memory leak?)
In-Reply-To: <5003054F.8040202@konink.de>
References: <5003054F.8040202@konink.de>
Message-ID: <CALdZE=vdvVH9juwmkcWLJ=a7SjeS_bDiC_HauKnJiwB0n61QCA@mail.gmail.com>

You suggest there are more issues. What are they? Thanks!

- Martin

On Sun, Jul 15, 2012 at 8:00 PM, Stefan de Konink <stefan at konink.de> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
>
> Hi,
>
> We are using ZeroMQ 2.2 and are quite happy about it. We wish we could
> move to ZeroMQ 3 but some outstanding bugs regarding the PubSub
> implementation prevent this.
>
> One of our current mysteries is the enormous high amount of memory our
> pubsub uses. The clients on the other hand consume a very high virtual
> memory amount. (Think in order of 500MB+)
>
> It seems that once in a few weeks we end up with a:
>
> FATAL ERROR: OUT OF MEMORY (yqueue.hpp:54)
>
>
> Our source code is pretty simple - no mallocs from our side what so
> ever and every init is closed; Still if I look to the virtual memory
> and real memory of one of my clients I see:
>
> 146MB virtual
> 56256 real
>
> ...where is this memory spend?
>
>
> Stefan
>
>
>
>
> #include <stdio.h>
> #include <stdlib.h>
> #include <assert.h>
> #include <zmq.h>
>
> int main (int argc, char *argv[]) {
>     if (argc != 3) {
>         printf("%s [receiver] [pubsub]\n\nEx.\n%s tcp://127.0.0.1:7807
> tcp://127.0.0.1:7817\n", argv[0], argv[0]);
>         exit(-1);
>     }
>
>         void *context  = zmq_init (1);
>         void *pubsub   = zmq_socket (context, ZMQ_PUB);
>         void *receiver = zmq_socket (context, ZMQ_PULL);
>
>         zmq_bind (pubsub,   argv[2]);
>         zmq_bind (receiver, argv[1]);
>
>         while (1) {
>                 int64_t more;
>                 size_t more_size = sizeof more;
>                 do {
>                         /* Create an empty 0MQ message to hold the
> message part */
>                         zmq_msg_t part;
>                         int rc = zmq_msg_init (&part);
>                         assert (rc == 0);
>                         /* Block until a message is available to be
> received from socket */
>                         rc = zmq_recv (receiver, &part, 0);
>                         assert (rc == 0);
>                         /* Determine if more message parts are to
> follow */
>                         rc = zmq_getsockopt (receiver, ZMQ_RCVMORE,
> &more, &more_size);
>                         assert (rc == 0);
>                         if (more) {
>                                 zmq_send (pubsub, &part, ZMQ_SNDMORE);
>                         } else {
>                                 zmq_send (pubsub, &part, 0);
>                         }
>                         zmq_msg_close (&part);
>                 } while (more);
>         }
> }
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.18 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
> iEYEAREKAAYFAlADBU4ACgkQYH1+F2Rqwn3iVwCeLZGRgTRVVHXVAL9smKLYnEbC
> +LkAoIcWRhK1YylyLUlpkDEtoLgF0qiS
> =Hy+T
> -----END PGP SIGNATURE-----
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From stefan at konink.de  Mon Jul 16 00:28:21 2012
From: stefan at konink.de (Stefan de Konink)
Date: Mon, 16 Jul 2012 00:28:21 +0200
Subject: [zeromq-dev] High memory usage (memory leak?)
In-Reply-To: <CALdZE=vdvVH9juwmkcWLJ=a7SjeS_bDiC_HauKnJiwB0n61QCA@mail.gmail.com>
References: <5003054F.8040202@konink.de>
	<CALdZE=vdvVH9juwmkcWLJ=a7SjeS_bDiC_HauKnJiwB0n61QCA@mail.gmail.com>
Message-ID: <50034405.7010806@konink.de>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 16-07-12 00:10, Martin Hurton wrote:
> You suggest there are more issues. What are they? Thanks!

Our migration issue before moving to ZeroMQ3 is:

<https://zeromq.jira.com/browse/LIBZMQ-248>

It seems to impossible to receive *any* message for some reason by the
client using the exact same programs (obviously the C code is
modified). While we mainly use PyZMQ, the same issue seems to be with
native C code. I suspect it has something to do with multipart messages.


But the memory issue really gives a headache, so kinda have priority
to get addressed.


Stefan
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.18 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEAREKAAYFAlADRAUACgkQYH1+F2Rqwn0MMQCdFOos8wSfFCDfXrM/+9LSE7py
3AgAnAy1JPeZq41axVeGcTaEGmxCtwmI
=ZG7q
-----END PGP SIGNATURE-----

From hurtonm at gmail.com  Mon Jul 16 00:48:31 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Mon, 16 Jul 2012 00:48:31 +0200
Subject: [zeromq-dev] High memory usage (memory leak?)
In-Reply-To: <50034405.7010806@konink.de>
References: <5003054F.8040202@konink.de>
	<CALdZE=vdvVH9juwmkcWLJ=a7SjeS_bDiC_HauKnJiwB0n61QCA@mail.gmail.com>
	<50034405.7010806@konink.de>
Message-ID: <CALdZE=vvUeeCh8MEXXiSSrORLA8Z6R2zD+CWPT+uc8R2Gq7t=A@mail.gmail.com>

Stefan, could you please create a Jira issue for this?

- Martin

On Mon, Jul 16, 2012 at 12:28 AM, Stefan de Konink <stefan at konink.de> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
>
> On 16-07-12 00:10, Martin Hurton wrote:
>> You suggest there are more issues. What are they? Thanks!
>
> Our migration issue before moving to ZeroMQ3 is:
>
> <https://zeromq.jira.com/browse/LIBZMQ-248>
>
> It seems to impossible to receive *any* message for some reason by the
> client using the exact same programs (obviously the C code is
> modified). While we mainly use PyZMQ, the same issue seems to be with
> native C code. I suspect it has something to do with multipart messages.
>
>
> But the memory issue really gives a headache, so kinda have priority
> to get addressed.
>
>
> Stefan
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.18 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
> iEYEAREKAAYFAlADRAUACgkQYH1+F2Rqwn0MMQCdFOos8wSfFCDfXrM/+9LSE7py
> 3AgAnAy1JPeZq41axVeGcTaEGmxCtwmI
> =ZG7q
> -----END PGP SIGNATURE-----
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From Zafar.Iqbal at lnties.com  Mon Jul 16 02:27:00 2012
From: Zafar.Iqbal at lnties.com (Zafar Iqbal)
Date: Mon, 16 Jul 2012 00:27:00 +0000
Subject: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX
In-Reply-To: <50007F49.8090502@steinhoff.de>
References: <7E56B0710035754D80A22F7BA63062C806DA7748@POCITMSEXMB01.LntUniverse.com>,
	<50002273.2060402@steinhoff.de>
	<7E56B0710035754D80A22F7BA63062C806DA777B@POCITMSEXMB01.LntUniverse.com>,
	<50007F49.8090502@steinhoff.de>
Message-ID: <7E56B0710035754D80A22F7BA63062C806DA77F2@POCITMSEXMB01.LntUniverse.com>

Thanks a lot, This Has Solved my problem,
Even zeroMQ-2.1.9 is also got compiled and now it is working fine in QNX.


________________________________________
From: Armin Steinhoff [armin at steinhoff.de]
Sent: Saturday, July 14, 2012 1:34 AM
To: Zafar Iqbal
Cc: ZeroMQ development list
Subject: Re: [zeromq-dev] Problem When Compiling Zeromq 2.1.9 in QNX

Hi,

I have recompiled zmq 2.2 and configured with

CC=qcc CXX=QCC ./configure
make
make install

without any problem for QNX 6.5 SP1 by using "bash".

BR

--Armin



Zafar Iqbal wrote:
>
>> Hi All,
>>
>> I am trying to compile Zeromq-2.1.9 in QNX self hosted system, I was getting the following errors.
>>
>> 1. Initially after running the configuration file, Makefile generated sucessfully. When I invoked make , the compilation started and when it reached to the linking stage it has given the error -lcrypto not found, just to chek it I removed -lcrypto from the makefile then it has shown the error -lsocket not found, When I looked into the libs path I found both the were available.
>    .... and what is your LD_LIBRARY_PATH ? Does it include /usr/lib ?
>
> Yes my LD_LIBRARY_PATH includes /usr/lib
>
>> After some analysis I found -rpath was not set, I set it to /usr/lib, after that programme compiled successfully.
>    ... not sure if the -rpath option works with QCC front end! Are you
> using QNX65 ?
>
> Yes I am using QNX65.
>
>
>
> http://steinhoff-automation.com
>
>>
>> But after that I faced the Following problem:
>>
>> 2. When I tried to install it giving make install:
>> I got the following error.
>>
>>
>> Making install in src
>> make[1]: Entering directory `/root/zeromq-2.1.9/src'
>> make[2]: Entering directory `/root/zeromq-2.1.9/src'
>> test -z "/usr/local/lib" || ../config/install-sh -c -d "/usr/local/lib"
>>    /bin/sh ../libtool   --mode=install ../config/install-sh -c   libzmq.la '/usr/loca
>> libtool: install: ../config/install-sh -c .libs/libzmq.lai /usr/local/lib/libzmq.la
>> ../config/install-sh: .libs/libzmq.lai does not exist.
>> make[2]: *** [install-libLTLIBRARIES] Error 1
>> make[2]: Leaving directory `/root/zeromq-2.1.9/src'
>> make[1]: *** [install-am] Error 2
>> make[1]: Leaving directory `/root/zeromq-2.1.9/src'
>> make: *** [install-recursive] Error 1
>>
>>
>> I am not able to find out what is the actual problem, How to solve the same.
>>
>> Please let me know if any of you have gone through the similar kind of problem.
>>
>>
>>
>> Thanking a lot.
>>
>> Larsen & Toubro Limited
>>
>> www.larsentoubro.com
>>
>> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
> Larsen & Toubro Limited
>
> www.larsentoubro.com
>
> This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.
>


Larsen & Toubro Limited

www.larsentoubro.com

This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.

From Zafar.Iqbal at lnties.com  Mon Jul 16 02:39:21 2012
From: Zafar.Iqbal at lnties.com (Zafar Iqbal)
Date: Mon, 16 Jul 2012 00:39:21 +0000
Subject: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 13
In-Reply-To: <mailman.1.1342346402.16522.zeromq-dev@lists.zeromq.org>
References: <mailman.1.1342346402.16522.zeromq-dev@lists.zeromq.org>
Message-ID: <7E56B0710035754D80A22F7BA63062C806DA7804@POCITMSEXMB01.LntUniverse.com>

HI,

Download third party ISO from the following link:

http://www.qnx.com/download/feature.html?programid=9978

mount the iso image to your filesystem(if you are using Linux or QNX as self host ) extract src.zip from the image you will find bash folder in it.

run the command CC=qcc  ./configure followed by make and make install.
Larsen & Toubro Limited

www.larsentoubro.com

This Email may contain confidential or privileged information for the intended recipient (s). If you are not the intended recipient, please do not use or disseminate the information, notify the sender and delete it from your system.

From victor at iso3103.net  Mon Jul 16 08:29:59 2012
From: victor at iso3103.net (Victor Perron)
Date: Mon, 16 Jul 2012 08:29:59 +0200
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B384D505B@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B384D505B@ex10czdmb01>
Message-ID: <CAF9FejFv=yNcRnHdUA7dOhKuNXxQ8Bjrpzx=bZ_EqtLm-QF5Mg@mail.gmail.com>

Hello,

Did you properly follow the directions stated at
http://www.zeromq.org/build:android ?
According to those, you shouldn't even have generated two binaries in the
first place.
Renaming libzmq.so.3.0.0 into libzmq.so has also almost no chance to work
the way you want it to work; it does not fix the actual soname which is
hardcoded to "libzmq.so.3.0.0" in both libs at link time.

I'd advise you to follow the steps at that url, and report any further
issue you'd have !

Best regards,

On Sun, Jul 15, 2012 at 2:38 PM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com>wrote:

>  Hello,****
>
> ** **
>
> I managed to compile the ZeroMQ + Java binding shared objects for ARM.****
>
> Thias is what file gives me on the two so files ****
>
> libzmq.so.3.0.0: ELF 32-bit LSB shared object, ARM, version 1 (SYSV),
> dynamically linked, not stripped****
>
> libjzmq.so: ELF 32-bit LSB shared object, ARM, version 1 (SYSV),
> dynamically linked, not stripped****
>
> and of course, I also have zmq.jar****
>
> now I created a new Android project in Eclipse.****
>
> Following advice on the net:****
>
> **?        **I placed the two so files udner libs/armeabi (renamed
> libzmq.so.3.0.0 into libzmq.so)****
>
> **?        **I placed the jar file under libs, and added to the build path
> ****
>
> **?        **Took the tripping sample and called it from main activity****
>
> **?        **Compile was successful. However, debugging the apk on a
> mobile device, I get the following error during System.loadLibrary in the
> ctor of ZMQ
> file name: /data/data/com.avgmobilation.zeromq/lib/libjzmq.so
> Cannot load library: reloc_library[1289]:   157 cannot locate
> '_ZNSs6assignEPKcj'...
> and later on this gets translated into an UnsatisfiedLinkError****
>
> ** **
>
> Even tried to build the apk on the linux machine where the so was compiled
> ? same error.****
>
> Any help is appreciated?****
>
> ** **
>
> Thanks,****
>
> ** **
>
> Sharon****
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120716/7f00e9dd/attachment.htm 

From aalok at ideadevice.com  Mon Jul 16 12:17:41 2012
From: aalok at ideadevice.com (aalok sood)
Date: Mon, 16 Jul 2012 15:47:41 +0530
Subject: [zeromq-dev] A query about implementing aborts in request reply
	broker
In-Reply-To: <CAEYnEwE6-VwvMyikB+PFeTB9kS_mNf5qLgubKL0iy6eq0Q+05Q@mail.gmail.com>
References: <CAG+SKy5t68xpGrgFdFoWoz9SjDZv9XjuK8A0-TX8OYTGStbdtg@mail.gmail.com>
	<CAEYnEwE6-VwvMyikB+PFeTB9kS_mNf5qLgubKL0iy6eq0Q+05Q@mail.gmail.com>
Message-ID: <CAG+SKy6ybC3esb=ob-Evefxj1ZS0AkCQssk8sdmEO8scq3NE1Q@mail.gmail.com>

Thanks for your suggestion Ian.

I have a similar model in place, but the issue is how do I find out how
many responses are going to come in on my pull socket.

In essence, can I somehow find out how many backend workers are connected
to my broker.


Regards
Aalok
On Sun, Jul 15, 2012 at 9:10 PM, Ian Barber <ian.barber at gmail.com> wrote:

> On Sat, Jul 14, 2012 at 7:52 PM, aalok sood <aalok at ideadevice.com> wrote:
>
>> I have a request reply broker and it does the job really well.
>>
>> Now if I want to tell my workers to abort a bunch of these requests,
>> based on some request parameter.
>>
>> For this I can use a pub-sub, but how do I collect the results?
>>
>
> I would use a pub-sub and a pull or similar socket - so requests to cancel
> are published, and you get a confirm message when they've been cancelled
> sent back over the pull socket. That said, given you req-rep at the moment,
> I'm guessing your work distributor is a dealer or router socket, so you
> could collect the results directly through that.
>
> Ian
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120716/6a91e94a/attachment.htm 

From markostrajkov at gmail.com  Mon Jul 16 18:38:12 2012
From: markostrajkov at gmail.com (Marko Trajkov)
Date: Mon, 16 Jul 2012 09:38:12 -0700
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
Message-ID: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>

Hello,

I am having trouble in running simple Hello World example in Perl. I
installed ZeroMQ and perl module, but when I start perl scripts, they are
running without errors but client and server does not exchange messages (do
not receive messages and does not sends it) everything else is working
without errors. This are example which I am trying to run.

https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl

When client sends "Hello" tcpdump doesn't catch it, so I can conclude that
package was not sent. When I manually send package using sendip, server
doesn't received it.

Please tell me if there is something that I was missing, where I am wrong.
I don't know how to make it working. By the way, I am using zeroMq 2.2,
ubuntu linux 12.04 and Perl 5.14.2

Second question: Is there UDP protocol for zeroMQ sockets?

Kind Regards,

Marko Trajkov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120716/0f09fe8f/attachment.htm 

From pelletier.michel at gmail.com  Mon Jul 16 19:02:21 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 16 Jul 2012 10:02:21 -0700
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
In-Reply-To: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>
References: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>
Message-ID: <CACxu=vKAkcBDN0G23XtFdY9rBom+HE8UrJWaHEHFfrn0NPatbw@mail.gmail.com>

Your client and server are not connected to each other.  They are both
calling connect() and being passed two different URIs.

One of your components has to call bind (which one is up to you, but
probably your server) and bind to an address, then the other component
must call connect to connect to that address.  Please read the guide
which goes into specific details about this right in the beginning.

http://zguide.zeromq.org/page:all

-Michel

On Mon, Jul 16, 2012 at 9:38 AM, Marko Trajkov <markostrajkov at gmail.com> wrote:
> Hello,
>
> I am having trouble in running simple Hello World example in Perl. I
> installed ZeroMQ and perl module, but when I start perl scripts, they are
> running without errors but client and server does not exchange messages (do
> not receive messages and does not sends it) everything else is working
> without errors. This are example which I am trying to run.
>
> https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
> https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl
>
> When client sends "Hello" tcpdump doesn't catch it, so I can conclude that
> package was not sent. When I manually send package using sendip, server
> doesn't received it.
>
> Please tell me if there is something that I was missing, where I am wrong. I
> don't know how to make it working. By the way, I am using zeroMq 2.2, ubuntu
> linux 12.04 and Perl 5.14.2
>
> Second question: Is there UDP protocol for zeroMQ sockets?
>
> Kind Regards,
>
> Marko Trajkov
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From pelletier.michel at gmail.com  Mon Jul 16 19:07:19 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 16 Jul 2012 10:07:19 -0700
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
In-Reply-To: <CACxu=vKAkcBDN0G23XtFdY9rBom+HE8UrJWaHEHFfrn0NPatbw@mail.gmail.com>
References: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>
	<CACxu=vKAkcBDN0G23XtFdY9rBom+HE8UrJWaHEHFfrn0NPatbw@mail.gmail.com>
Message-ID: <CACxu=vJOPNm5sV4SmQ4um721VtgZk6tdPdepQ87b_=NasAC8dQ@mail.gmail.com>

Ah I see you got those examples *from* the guide, I would say those
are broken.  Your best bet is to have your server bind() on a
localhost address and your client connect().  Check out the equivalent
 code in Python:

https://github.com/imatix/zguide/blob/master/examples/Python/rrserver.py

-Michel

On Mon, Jul 16, 2012 at 10:02 AM, Michel Pelletier
<pelletier.michel at gmail.com> wrote:
> Your client and server are not connected to each other.  They are both
> calling connect() and being passed two different URIs.
>
> One of your components has to call bind (which one is up to you, but
> probably your server) and bind to an address, then the other component
> must call connect to connect to that address.  Please read the guide
> which goes into specific details about this right in the beginning.
>
> http://zguide.zeromq.org/page:all
>
> -Michel
>
> On Mon, Jul 16, 2012 at 9:38 AM, Marko Trajkov <markostrajkov at gmail.com> wrote:
>> Hello,
>>
>> I am having trouble in running simple Hello World example in Perl. I
>> installed ZeroMQ and perl module, but when I start perl scripts, they are
>> running without errors but client and server does not exchange messages (do
>> not receive messages and does not sends it) everything else is working
>> without errors. This are example which I am trying to run.
>>
>> https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
>> https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl
>>
>> When client sends "Hello" tcpdump doesn't catch it, so I can conclude that
>> package was not sent. When I manually send package using sendip, server
>> doesn't received it.
>>
>> Please tell me if there is something that I was missing, where I am wrong. I
>> don't know how to make it working. By the way, I am using zeroMq 2.2, ubuntu
>> linux 12.04 and Perl 5.14.2
>>
>> Second question: Is there UDP protocol for zeroMQ sockets?
>>
>> Kind Regards,
>>
>> Marko Trajkov
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>

From abc at alexsergeyev.com  Tue Jul 17 03:55:51 2012
From: abc at alexsergeyev.com (Alex Sergeyev)
Date: Mon, 16 Jul 2012 21:55:51 -0400
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
In-Reply-To: <CACxu=vJOPNm5sV4SmQ4um721VtgZk6tdPdepQ87b_=NasAC8dQ@mail.gmail.com>
References: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>
	<CACxu=vKAkcBDN0G23XtFdY9rBom+HE8UrJWaHEHFfrn0NPatbw@mail.gmail.com>
	<CACxu=vJOPNm5sV4SmQ4um721VtgZk6tdPdepQ87b_=NasAC8dQ@mail.gmail.com>
Message-ID: <CAJbypPr4GJ=FRevqjWZ3OAM1OZSA4FwQx+fqNwxcnSDd9gJA0Q@mail.gmail.com>

This would be appropriate to commit:

------------------------------------------------
--- a/examples/Perl/rrserver.pl
+++ b/examples/Perl/rrserver.pl
@@ -21,7 +21,7 @@ my $context = ZeroMQ::Context->new();

 # Socket to talk to clients
 my $responder = $context->socket(ZMQ_REP);
-$responder->connect('tcp://localhost:5560');
+$responder->bind('tcp://*:5559');

 while (1) {
     # Wait for next request from client
------------------------------------------------

(localhost is ambiguous IPv6 vs IPv4, so I used * for simplicity)

From anti_tenzor at mail.ru  Tue Jul 17 09:37:51 2012
From: anti_tenzor at mail.ru (anti_tenzor)
Date: Tue, 17 Jul 2012 11:37:51 +0400
Subject: [zeromq-dev] what are the best practices to send complex objects?
Message-ID: <5005164F.9030603@mail.ru>

Hello, all,

I'd like to use ZMQ to send complex objects between applications (using 
CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
I am forced to implement some serialization-deserialization mechanism.

But the hole idea is to be as fast as possible! So it looks like this 
algo should be:
1. Extremely fast
2. Produce very compact byte arrays

Could you please suggest some best practices or libraries (.net), that 
meets these requirements?

How do you send complex ojects via ZMQ in real-life applications?

Thank you in advance!


From ronaldbelliott at gmail.com  Tue Jul 17 09:40:01 2012
From: ronaldbelliott at gmail.com (Ron Elliott)
Date: Tue, 17 Jul 2012 00:40:01 -0700
Subject: [zeromq-dev] what are the best practices to send complex
	objects?
In-Reply-To: <5005164F.9030603@mail.ru>
References: <5005164F.9030603@mail.ru>
Message-ID: <CAMEND1j9OEYWp5Q_YZAGvwoDBuOfAYQH6xPE6gUFSS0kPJfAJg@mail.gmail.com>

Not sure about .NET compatibility but check out Google's Protocol Buffers
http://code.google.com/p/protobuf/
On Jul 17, 2012 12:38 AM, "anti_tenzor" <anti_tenzor at mail.ru> wrote:

> Hello, all,
>
> I'd like to use ZMQ to send complex objects between applications (using
> CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
> I am forced to implement some serialization-deserialization mechanism.
>
> But the hole idea is to be as fast as possible! So it looks like this
> algo should be:
> 1. Extremely fast
> 2. Produce very compact byte arrays
>
> Could you please suggest some best practices or libraries (.net), that
> meets these requirements?
>
> How do you send complex ojects via ZMQ in real-life applications?
>
> Thank you in advance!
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/c0af5b09/attachment.htm 

From aalok at ideadevice.com  Tue Jul 17 09:40:21 2012
From: aalok at ideadevice.com (aalok sood)
Date: Tue, 17 Jul 2012 13:10:21 +0530
Subject: [zeromq-dev] what are the best practices to send complex
	objects?
In-Reply-To: <5005164F.9030603@mail.ru>
References: <5005164F.9030603@mail.ru>
Message-ID: <CAG+SKy7D2wstq2-mC--PLzt6j7FnzHue-8yoH82iyz0xXoG9Vw@mail.gmail.com>

You could look at json. I have been using it quite heavily
for serialization-deserialization, and it does the job quite well.

On Tue, Jul 17, 2012 at 1:07 PM, anti_tenzor <anti_tenzor at mail.ru> wrote:

> Hello, all,
>
> I'd like to use ZMQ to send complex objects between applications (using
> CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
> I am forced to implement some serialization-deserialization mechanism.
>
> But the hole idea is to be as fast as possible! So it looks like this
> algo should be:
> 1. Extremely fast
> 2. Produce very compact byte arrays
>
> Could you please suggest some best practices or libraries (.net), that
> meets these requirements?
>
> How do you send complex ojects via ZMQ in real-life applications?
>
> Thank you in advance!
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/8303ecdd/attachment.htm 

From Sharon.Ben-Asher at avg.com  Tue Jul 17 10:09:39 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Tue, 17 Jul 2012 08:09:39 +0000
Subject: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 14
In-Reply-To: <mailman.1.1342432801.30315.zeromq-dev@lists.zeromq.org>
References: <mailman.1.1342432801.30315.zeromq-dev@lists.zeromq.org>
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D638A@ex10czdmb01>

Hi Victor,

I did follow the instructions in the Build Android wiki.
I don't know what you mean by "you shouldn't even have generated two binaries in the first place " - I have the *exact* directory tree that is described in that page for ZeroMQ 3.x. 
The listed directory tree shows two shared object files: libzmq and libjzmq.
The instructions stop after production of zmq.jar, there are no instructions on how to embed the product into an APK, which .so files to put there, etc.
I tried to put only libjzmq.so in the apk (under libs/armeabi) with the same error.
I even enlisted our expert APK developer, he tried several configurations and file placements, with no avail.

I would like to know which steps you think I did not follow, or otherwise what am I missing (no sarcasm intended, I really am at a loss here)

Thanks,

Sharon.


-----Original Message-----
From: Victor Perron <victor at iso3103.net>
Subject: Re: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org>
Message-ID:
	<CAF9FejFv=yNcRnHdUA7dOhKuNXxQ8Bjrpzx=bZ_EqtLm-QF5Mg at mail.gmail.com>
Content-Type: text/plain; charset="windows-1252"

Hello,

Did you properly follow the directions stated at http://www.zeromq.org/build:android ?
According to those, you shouldn't even have generated two binaries in the first place.
Renaming libzmq.so.3.0.0 into libzmq.so has also almost no chance to work the way you want it to work; it does not fix the actual soname which is hardcoded to "libzmq.so.3.0.0" in both libs at link time.

I'd advise you to follow the steps at that url, and report any further issue you'd have !

Best regards,

On Sun, Jul 15, 2012 at 2:38 PM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com>wrote:

>  Hello,****
>
> ** **
>
> I managed to compile the ZeroMQ + Java binding shared objects for 
> ARM.****
>
> Thias is what file gives me on the two so files ****
>
> libzmq.so.3.0.0: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), 
> dynamically linked, not stripped****
>
> libjzmq.so: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), 
> dynamically linked, not stripped****
>
> and of course, I also have zmq.jar****
>
> now I created a new Android project in Eclipse.****
>
> Following advice on the net:****
>
> **?        **I placed the two so files udner libs/armeabi (renamed
> libzmq.so.3.0.0 into libzmq.so)****
>
> **?        **I placed the jar file under libs, and added to the build path
> ****
>
> **?        **Took the tripping sample and called it from main activity****
>
> **?        **Compile was successful. However, debugging the apk on a
> mobile device, I get the following error during System.loadLibrary in 
> the ctor of ZMQ file name: 
> /data/data/com.avgmobilation.zeromq/lib/libjzmq.so
> Cannot load library: reloc_library[1289]:   157 cannot locate
> '_ZNSs6assignEPKcj'...
> and later on this gets translated into an UnsatisfiedLinkError****
>
> ** **
>
> Even tried to build the apk on the linux machine where the so was 
> compiled ? same error.****
>
> Any help is appreciated?****
>
> ** **
>
> Thanks,****
>
> ** **
>
> Sharon****
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


--
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120716/7f00e9dd/attachment-0001.htm 

------------------------------

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


End of zeromq-dev Digest, Vol 55, Issue 14
******************************************

From victor at iso3103.net  Tue Jul 17 10:38:09 2012
From: victor at iso3103.net (Victor Perron)
Date: Tue, 17 Jul 2012 10:38:09 +0200
Subject: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 14
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B384D638A@ex10czdmb01>
References: <mailman.1.1342432801.30315.zeromq-dev@lists.zeromq.org>
	<671E27F33F2F25499F78A4AFA494E63B384D638A@ex10czdmb01>
Message-ID: <CAF9FejHK-9LOrwhHMrinkeaiq5AkwAcxijVV9K6X45j5d__EaQ@mail.gmail.com>

Hi,

In that case it seems that you discovered something serious.
I took a look at it, and indeed the assign() symbol from std::string class
seems undefined.
That would mean, Android does _not_ define such a function in its SDK.

Now, the proper way to get around this is a patch to Zeromq 3.x.
You'll have to fix those lines (my guess):

./src/ipc_listener.cpp:151:    filename.assign(addr_);
./src/socket_base.cpp:331:            options.last_endpoint.assign (addr_);
./src/socket_base.cpp:480:        options.last_endpoint.assign (addr_);
./src/tcp_address.cpp:503:        addr_str.assign (name_, delimiter -
name_);
./src/tcp_address.cpp:504:        mask_str.assign (delimiter + 1);
./src/tcp_address.cpp:511:        addr_str.assign (name_);

... and try to compile again. If there are no other undefined symbols, that
should work.
That issue did not exist in ZeroMQ 2.2 though, that one didn't make an
extensive use of std::string.

Keep me updated, we'll see for a pull request when I'll be less busy ;)

Best regards,

On Tue, Jul 17, 2012 at 10:09 AM, Sharon Ben-Asher <Sharon.Ben-Asher at avg.com
> wrote:

> Hi Victor,
>
> I did follow the instructions in the Build Android wiki.
> I don't know what you mean by "you shouldn't even have generated two
> binaries in the first place " - I have the *exact* directory tree that is
> described in that page for ZeroMQ 3.x.
> The listed directory tree shows two shared object files: libzmq and
> libjzmq.
> The instructions stop after production of zmq.jar, there are no
> instructions on how to embed the product into an APK, which .so files to
> put there, etc.
> I tried to put only libjzmq.so in the apk (under libs/armeabi) with the
> same error.
> I even enlisted our expert APK developer, he tried several configurations
> and file placements, with no avail.
>
> I would like to know which steps you think I did not follow, or otherwise
> what am I missing (no sarcasm intended, I really am at a loss here)
>
> Thanks,
>
> Sharon.
>
>
> -----Original Message-----
> From: Victor Perron <victor at iso3103.net>
> Subject: Re: [zeromq-dev] ZeroMQ for Android: Problem in building the
>         Android APK
> To: ZeroMQ development list <zeromq-dev at lists.zeromq.org>
> Message-ID:
>         <CAF9FejFv=yNcRnHdUA7dOhKuNXxQ8Bjrpzx=
> bZ_EqtLm-QF5Mg at mail.gmail.com>
> Content-Type: text/plain; charset="windows-1252"
>
> Hello,
>
> Did you properly follow the directions stated at
> http://www.zeromq.org/build:android ?
> According to those, you shouldn't even have generated two binaries in the
> first place.
> Renaming libzmq.so.3.0.0 into libzmq.so has also almost no chance to work
> the way you want it to work; it does not fix the actual soname which is
> hardcoded to "libzmq.so.3.0.0" in both libs at link time.
>
> I'd advise you to follow the steps at that url, and report any further
> issue you'd have !
>
> Best regards,
>
> On Sun, Jul 15, 2012 at 2:38 PM, Sharon Ben-Asher
> <Sharon.Ben-Asher at avg.com>wrote:
>
> >  Hello,****
> >
> > ** **
> >
> > I managed to compile the ZeroMQ + Java binding shared objects for
> > ARM.****
> >
> > Thias is what file gives me on the two so files ****
> >
> > libzmq.so.3.0.0: ELF 32-bit LSB shared object, ARM, version 1 (SYSV),
> > dynamically linked, not stripped****
> >
> > libjzmq.so: ELF 32-bit LSB shared object, ARM, version 1 (SYSV),
> > dynamically linked, not stripped****
> >
> > and of course, I also have zmq.jar****
> >
> > now I created a new Android project in Eclipse.****
> >
> > Following advice on the net:****
> >
> > **?        **I placed the two so files udner libs/armeabi (renamed
> > libzmq.so.3.0.0 into libzmq.so)****
> >
> > **?        **I placed the jar file under libs, and added to the build
> path
> > ****
> >
> > **?        **Took the tripping sample and called it from main
> activity****
> >
> > **?        **Compile was successful. However, debugging the apk on a
> > mobile device, I get the following error during System.loadLibrary in
> > the ctor of ZMQ file name:
> > /data/data/com.avgmobilation.zeromq/lib/libjzmq.so
> > Cannot load library: reloc_library[1289]:   157 cannot locate
> > '_ZNSs6assignEPKcj'...
> > and later on this gets translated into an UnsatisfiedLinkError****
> >
> > ** **
> >
> > Even tried to build the apk on the linux machine where the so was
> > compiled ? same error.****
> >
> > Any help is appreciated?****
> >
> > ** **
> >
> > Thanks,****
> >
> > ** **
> >
> > Sharon****
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
> >
>
>
> --
> Victor
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120716/7f00e9dd/attachment-0001.htm
>
> ------------------------------
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
> End of zeromq-dev Digest, Vol 55, Issue 14
> ******************************************
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/ce2dc6db/attachment.htm 

From lestrrat at gmail.com  Tue Jul 17 10:57:29 2012
From: lestrrat at gmail.com (Daisuke Maki)
Date: Tue, 17 Jul 2012 17:57:29 +0900
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
In-Reply-To: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>
References: <CA+dFAu15FNK7BtpXYe2VRhxjuYVFgg8Y9S9oFyur0tjdiKCoqA@mail.gmail.com>
Message-ID: <CAK_8zA44H4L7VEC6o6+C9WWtHiR9srF4J=TtJvMQdZ6sizy39Q@mail.gmail.com>

Hi,

FYI I have gone on to restructure the Perl bindings like so:

   https://github.com/lestrrat/p5-ZMQ/
   # they are also available on CPAN

basically, I moved out the version specific stuff out to ZMQ::LibZMQ2
and ZMQ::LibZMQ3 so that I don't have to put #ifdef s all over the
place.

porting from ZeroMQ to ZMQ::LibZMQ* should be pretty straight forward.
Please let me know if you have problems at
https://github.com/lestrrat/p5-ZMQ/issues

--d

2012/7/17 Marko Trajkov <markostrajkov at gmail.com>:
> Hello,
>
> I am having trouble in running simple Hello World example in Perl. I
> installed ZeroMQ and perl module, but when I start perl scripts, they are
> running without errors but client and server does not exchange messages (do
> not receive messages and does not sends it) everything else is working
> without errors. This are example which I am trying to run.
>
> https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
> https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl
>
> When client sends "Hello" tcpdump doesn't catch it, so I can conclude that
> package was not sent. When I manually send package using sendip, server
> doesn't received it.
>
> Please tell me if there is something that I was missing, where I am wrong. I
> don't know how to make it working. By the way, I am using zeroMq 2.2, ubuntu
> linux 12.04 and Perl 5.14.2
>
> Second question: Is there UDP protocol for zeroMQ sockets?
>
> Kind Regards,
>
> Marko Trajkov
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From cfkaran2 at gmail.com  Tue Jul 17 11:57:06 2012
From: cfkaran2 at gmail.com (Cem Karan)
Date: Tue, 17 Jul 2012 05:57:06 -0400
Subject: [zeromq-dev] what are the best practices to send complex
	objects?
In-Reply-To: <5005164F.9030603@mail.ru>
References: <5005164F.9030603@mail.ru>
Message-ID: <94236CF4-F8CF-492A-AD4C-1A3653FC107B@gmail.com>

I've been playing with MessagePack (http://msgpack.org/) for serialization and BLOSC (http://blosc.pytables.org/trac) for compression.  They seem to work pretty well together.  How fast is fast?  BLOSC is designed to move data around in memory faster than memcpy() is able to do it, and I haven't had any complaints about how fast MessagePack works (OTOH, I'm I/O bound, so I tend not to hit any limits).

Thanks,
Cem Karan

On Jul 17, 2012, at 3:37 AM, anti_tenzor wrote:

> Hello, all,
> 
> I'd like to use ZMQ to send complex objects between applications (using 
> CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
> I am forced to implement some serialization-deserialization mechanism.
> 
> But the hole idea is to be as fast as possible! So it looks like this 
> algo should be:
> 1. Extremely fast
> 2. Produce very compact byte arrays
> 
> Could you please suggest some best practices or libraries (.net), that 
> meets these requirements?
> 
> How do you send complex ojects via ZMQ in real-life applications?
> 
> Thank you in advance!
> 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1581 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/513e2126/attachment.bin 

From aalok at ideadevice.com  Tue Jul 17 12:01:13 2012
From: aalok at ideadevice.com (aalok sood)
Date: Tue, 17 Jul 2012 15:31:13 +0530
Subject: [zeromq-dev] what are the best practices to send complex
	objects?
In-Reply-To: <94236CF4-F8CF-492A-AD4C-1A3653FC107B@gmail.com>
References: <5005164F.9030603@mail.ru>
	<94236CF4-F8CF-492A-AD4C-1A3653FC107B@gmail.com>
Message-ID: <CAG+SKy6jivviwWPHPmi3PrhZX3aP7JdTxrRwTZSddocXHK7ibw@mail.gmail.com>

If you are **really** concerned about performance, I suggest you have your
own protocol that both sides understand.
This way you send data as it is, the remote side would know how to
interpret it.



On Tue, Jul 17, 2012 at 3:27 PM, Cem Karan <cfkaran2 at gmail.com> wrote:

> I've been playing with MessagePack (http://msgpack.org/) for
> serialization and BLOSC (http://blosc.pytables.org/trac) for compression.
>  They seem to work pretty well together.  How fast is fast?  BLOSC is
> designed to move data around in memory faster than memcpy() is able to do
> it, and I haven't had any complaints about how fast MessagePack works
> (OTOH, I'm I/O bound, so I tend not to hit any limits).
>
> Thanks,
> Cem Karan
>
> On Jul 17, 2012, at 3:37 AM, anti_tenzor wrote:
>
> > Hello, all,
> >
> > I'd like to use ZMQ to send complex objects between applications (using
> > CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
> > I am forced to implement some serialization-deserialization mechanism.
> >
> > But the hole idea is to be as fast as possible! So it looks like this
> > algo should be:
> > 1. Extremely fast
> > 2. Produce very compact byte arrays
> >
> > Could you please suggest some best practices or libraries (.net), that
> > meets these requirements?
> >
> > How do you send complex ojects via ZMQ in real-life applications?
> >
> > Thank you in advance!
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/ef6cb1f2/attachment.htm 

From dophinehk at yahoo.com.hk  Tue Jul 17 12:36:09 2012
From: dophinehk at yahoo.com.hk (Kira)
Date: Tue, 17 Jul 2012 18:36:09 +0800 (SGT)
Subject: [zeromq-dev] help Handling Multiple Sockets
Message-ID: <1342521369.54808.YahooMailNeo@web190802.mail.sg3.yahoo.com>

Hi,

I try to write a program to handle multiple sockets.

1.
typedef std::vector<zmq::pollitem_t> tPollItemList;
tPollItemList m_oPollItemList;

2. add one entry to m_oPollItemList.
3. call ?zmq_poll (&m_oPollItemList[0], m_oPollItemList.size(), -1);

I found revents is always zero i.e.
tItr->revents & ZMQ_POLLIN is always false.

Does anyone know the cause?

Regards,
Kira
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/5e688f5a/attachment.htm 

From mimir at newmail.ru  Tue Jul 17 13:27:34 2012
From: mimir at newmail.ru (moteus)
Date: Tue, 17 Jul 2012 11:27:34 +0000 (UTC)
Subject: [zeromq-dev] Socket use from several threads
Message-ID: <loom.20120717T130445-42@post.gmane.org>

Hi.

I have some service. 
Service creates new thread for each request and call function from .dll to
proceed this request.
I write only .dll and can not change service.
What best method to use REQ/REP sockets over inproc and tcp.
If I understood correctly i have to create new context (tcp) and socket for each
new thread? Or i can create some pool of sockets and use them at the same time
only from one thread?


From markostrajkov at gmail.com  Tue Jul 17 14:17:56 2012
From: markostrajkov at gmail.com (Marko Trajkov)
Date: Tue, 17 Jul 2012 14:17:56 +0200
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
Message-ID: <CA+dFAu2GAR59rA7Yc2M7CXOfKmUtoSZ7fUN+UA4TYkhtaD=f9Q@mail.gmail.com>

Hi,

thanks for replays, I really appreciate it, but I still didn't have success
in making it working. Server binds on tcp://192.168.109.255:5560 (my local
address) and client connect to same address, bat they don't communicate.
When they try to receive data (recv()->data;) both becomes blocked.
In the example, client sends "Hello" but when I use tcpdump to monitor all
traffic, I don't see that anything was sent.

By the way, is it possible to use UDP sockets in ZeroMQ, I don't see that
something like this is possible in the guide, only (inproc, ipc, tcp, pgm,
epgm).

Kind Regards,

Marko Trajkov

On Mon, Jul 16, 2012 at 9:38 AM, Marko Trajkov <markostrajkov at gmail.com>
wrote:
> Hello,
>
> I am having trouble in running simple Hello World example in Perl. I
> installed ZeroMQ and perl module, but when I start perl scripts, they are
> running without errors but client and server does not exchange messages
(do
> not receive messages and does not sends it) everything else is working
> without errors. This are example which I am trying to run.
>
> https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
> https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl
>
> When client sends "Hello" tcpdump doesn't catch it, so I can conclude that
> package was not sent. When I manually send package using sendip, server
> doesn't received it.
>
> Please tell me if there is something that I was missing, where I am
wrong. I
> don't know how to make it working. By the way, I am using zeroMq 2.2,
ubuntu
> linux 12.04 and Perl 5.14.2
>
> Second question: Is there UDP protocol for zeroMQ sockets?
>
> Kind Regards,
>
> Marko Trajkov
>
> ______________________________
_________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/03e1bc97/attachment.htm 

From lists at chuckremes.com  Tue Jul 17 14:58:34 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Tue, 17 Jul 2012 07:58:34 -0500
Subject: [zeromq-dev] Socket use from several threads
In-Reply-To: <loom.20120717T130445-42@post.gmane.org>
References: <loom.20120717T130445-42@post.gmane.org>
Message-ID: <75FBA2FE-975A-4949-A1A3-86251B2E8AB2@chuckremes.com>

On Jul 17, 2012, at 6:27 AM, moteus wrote:

> Hi.
> 
> I have some service. 
> Service creates new thread for each request and call function from .dll to
> proceed this request.
> I write only .dll and can not change service.
> What best method to use REQ/REP sockets over inproc and tcp.
> If I understood correctly i have to create new context (tcp) and socket for each
> new thread? Or i can create some pool of sockets and use them at the same time
> only from one thread?

You should create only ONE context. The context can be passed to each thread; any socket you create in a thread should only be used from that thread.

I recommend that you read the guide and the FAQ. These ideas are discussed in both places.

cr


From lestrrat at gmail.com  Tue Jul 17 15:18:20 2012
From: lestrrat at gmail.com (Daisuke Maki)
Date: Tue, 17 Jul 2012 22:18:20 +0900
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
In-Reply-To: <CA+dFAu2GAR59rA7Yc2M7CXOfKmUtoSZ7fUN+UA4TYkhtaD=f9Q@mail.gmail.com>
References: <CA+dFAu2GAR59rA7Yc2M7CXOfKmUtoSZ7fUN+UA4TYkhtaD=f9Q@mail.gmail.com>
Message-ID: <CAK_8zA6XuJ7rYSuKkT4iG8n08WvNxwMB3ddu0k72DKDiWJUDHg@mail.gmail.com>

Here's a working example using ZMQ::LibZMQ2

https://gist.github.com/3129346

Maybe I wasn't explicit enough, but ZeroMQ.pm is done for. I'm no
longer going to maintain it. please use ZMQ::LibZMQ2 or ZMQ::LibZMQ3

--d

2012/7/17 Marko Trajkov <markostrajkov at gmail.com>:
> Hi,
>
> thanks for replays, I really appreciate it, but I still didn't have success
> in making it working. Server binds on tcp://192.168.109.255:5560 (my local
> address) and client connect to same address, bat they don't communicate.
> When they try to receive data (recv()->data;) both becomes blocked.
> In the example, client sends "Hello" but when I use tcpdump to monitor all
> traffic, I don't see that anything was sent.
>
> By the way, is it possible to use UDP sockets in ZeroMQ, I don't see that
> something like this is possible in the guide, only (inproc, ipc, tcp, pgm,
> epgm).
>
> Kind Regards,
>
> Marko Trajkov
>
>
> On Mon, Jul 16, 2012 at 9:38 AM, Marko Trajkov <markostrajkov at gmail.com>
> wrote:
>> Hello,
>>
>> I am having trouble in running simple Hello World example in Perl. I
>> installed ZeroMQ and perl module, but when I start perl scripts, they are
>> running without errors but client and server does not exchange messages
>> (do
>> not receive messages and does not sends it) everything else is working
>> without errors. This are example which I am trying to run.
>>
>> https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
>> https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl
>>
>> When client sends "Hello" tcpdump doesn't catch it, so I can conclude that
>> package was not sent. When I manually send package using sendip, server
>> doesn't received it.
>>
>> Please tell me if there is something that I was missing, where I am wrong.
>> I
>> don't know how to make it working. By the way, I am using zeroMq 2.2,
>> ubuntu
>> linux 12.04 and Perl 5.14.2
>>
>> Second question: Is there UDP protocol for zeroMQ sockets?
>>
>> Kind Regards,
>>
>> Marko Trajkov
>>
>> ______________________________
> _________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From mimir at newmail.ru  Tue Jul 17 16:14:20 2012
From: mimir at newmail.ru (moteus)
Date: Tue, 17 Jul 2012 14:14:20 +0000 (UTC)
Subject: [zeromq-dev] Socket use from several threads
References: <loom.20120717T130445-42@post.gmane.org>
	<75FBA2FE-975A-4949-A1A3-86251B2E8AB2@chuckremes.com>
Message-ID: <loom.20120717T160608-968@post.gmane.org>

Chuck Remes <lists <at> chuckremes.com> writes:

> You should create only ONE context. The context can be passed to each thread;
any socket you create in a
> thread should only be used from that thread.
> 
> I recommend that you read the guide and the FAQ. These ideas are discussed in
both places.
> 
> cr
> 

I don't like to create a socket only for one request.
(connect=>send/recv=>close)
But as I understood there is no other way.
About context - sorry for a silly question :)






From lists at chuckremes.com  Tue Jul 17 16:26:51 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Tue, 17 Jul 2012 09:26:51 -0500
Subject: [zeromq-dev] Socket use from several threads
In-Reply-To: <loom.20120717T160608-968@post.gmane.org>
References: <loom.20120717T130445-42@post.gmane.org>
	<75FBA2FE-975A-4949-A1A3-86251B2E8AB2@chuckremes.com>
	<loom.20120717T160608-968@post.gmane.org>
Message-ID: <22DB4BFF-7566-4C54-B0A3-E7F0DDC54AA8@chuckremes.com>

On Jul 17, 2012, at 9:14 AM, moteus wrote:

> Chuck Remes <lists <at> chuckremes.com> writes:
> 
>> You should create only ONE context. The context can be passed to each thread;
> any socket you create in a
>> thread should only be used from that thread.
>> 
>> I recommend that you read the guide and the FAQ. These ideas are discussed in
> both places.
>> 
>> cr
>> 
> 
> I don't like to create a socket only for one request.
> (connect=>send/recv=>close)
> But as I understood there is no other way.
> About context - sorry for a silly question :)

Oh, so the threads are short-lived then. In that case, you should create a pool of them and have each thread "check out" the socket from the pool when it needs to use it and check it back in again when the thread exits. If the pool is protected by a mechanism (like a mutex) that employs a full memory barrier, then it will be safe.

cr


From markostrajkov at gmail.com  Tue Jul 17 16:34:45 2012
From: markostrajkov at gmail.com (Marko Trajkov)
Date: Tue, 17 Jul 2012 16:34:45 +0200
Subject: [zeromq-dev] ZeroMQ Hello World Perl Example Start Up
In-Reply-To: <CAK_8zA6XuJ7rYSuKkT4iG8n08WvNxwMB3ddu0k72DKDiWJUDHg@mail.gmail.com>
References: <CA+dFAu2GAR59rA7Yc2M7CXOfKmUtoSZ7fUN+UA4TYkhtaD=f9Q@mail.gmail.com>
	<CAK_8zA6XuJ7rYSuKkT4iG8n08WvNxwMB3ddu0k72DKDiWJUDHg@mail.gmail.com>
Message-ID: <CA+dFAu1DSk-wdF+GO_-4QLVsjnEu-vaQOwXu73YR4aduODZ4SQ@mail.gmail.com>

Thanks you. Now is working fine. I didn't understand you at first time.

Kind regards,
Trajkov Marko

On Tue, Jul 17, 2012 at 3:18 PM, Daisuke Maki <lestrrat at gmail.com> wrote:

> Here's a working example using ZMQ::LibZMQ2
>
> https://gist.github.com/3129346
>
> Maybe I wasn't explicit enough, but ZeroMQ.pm is done for. I'm no
> longer going to maintain it. please use ZMQ::LibZMQ2 or ZMQ::LibZMQ3
>
> --d
>
> 2012/7/17 Marko Trajkov <markostrajkov at gmail.com>:
> > Hi,
> >
> > thanks for replays, I really appreciate it, but I still didn't have
> success
> > in making it working. Server binds on tcp://192.168.109.255:5560 (my
> local
> > address) and client connect to same address, bat they don't communicate.
> > When they try to receive data (recv()->data;) both becomes blocked.
> > In the example, client sends "Hello" but when I use tcpdump to monitor
> all
> > traffic, I don't see that anything was sent.
> >
> > By the way, is it possible to use UDP sockets in ZeroMQ, I don't see that
> > something like this is possible in the guide, only (inproc, ipc, tcp,
> pgm,
> > epgm).
> >
> > Kind Regards,
> >
> > Marko Trajkov
> >
> >
> > On Mon, Jul 16, 2012 at 9:38 AM, Marko Trajkov <markostrajkov at gmail.com>
> > wrote:
> >> Hello,
> >>
> >> I am having trouble in running simple Hello World example in Perl. I
> >> installed ZeroMQ and perl module, but when I start perl scripts, they
> are
> >> running without errors but client and server does not exchange messages
> >> (do
> >> not receive messages and does not sends it) everything else is working
> >> without errors. This are example which I am trying to run.
> >>
> >> https://github.com/imatix/zguide/blob/master/examples/Perl/rrclient.pl
> >> https://github.com/imatix/zguide/blob/master/examples/Perl/rrserver.pl
> >>
> >> When client sends "Hello" tcpdump doesn't catch it, so I can conclude
> that
> >> package was not sent. When I manually send package using sendip, server
> >> doesn't received it.
> >>
> >> Please tell me if there is something that I was missing, where I am
> wrong.
> >> I
> >> don't know how to make it working. By the way, I am using zeroMq 2.2,
> >> ubuntu
> >> linux 12.04 and Perl 5.14.2
> >>
> >> Second question: Is there UDP protocol for zeroMQ sockets?
> >>
> >> Kind Regards,
> >>
> >> Marko Trajkov
> >>
> >> ______________________________
> > _________________
> >> zeromq-dev mailing list
> >> zeromq-dev at lists.zeromq.org
> >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >>
> >
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/17bb6128/attachment.htm 

From bluerocketdelivery at gmail.com  Tue Jul 17 08:22:04 2012
From: bluerocketdelivery at gmail.com (Ron Elliott)
Date: Mon, 16 Jul 2012 23:22:04 -0700
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
Message-ID: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>

Using the downloaded stable version 2.2 code I cannot get ZMQ to compile with the --with-pgm option. The project is too far along to make the switch to v3 so that's not an option atm. Any other solutions or is ZMQ dead for me?

Error:

Making all in build-staging/openpgm/pgm
...
In file included from ./include/impl/sockaddr.h:38,
                 from ./include/impl/notify.h:46,
                 from ./include/impl/framework.h:66,
                 from thread.c:23:
./include/pgm/in.h:34: error: redefinition of ?struct group_req?
./include/pgm/in.h:40: error: redefinition of ?struct group_source_req?
make[2]: *** [libpgm_noinst_la-thread.lo] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all-recursive] Error 1

Thanks in advance to anybody that may be able to provide me with some help!

From steven.mccoy at miru.hk  Tue Jul 17 21:27:17 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Tue, 17 Jul 2012 15:27:17 -0400
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
Message-ID: <CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>

On 17 July 2012 02:22, Ron Elliott <bluerocketdelivery at gmail.com> wrote:

> Using the downloaded stable version 2.2 code I cannot get ZMQ to compile
> with the --with-pgm option. The project is too far along to make the switch
> to v3 so that's not an option atm. Any other solutions or is ZMQ dead for
> me?
>
>
It's a known issue, I have to upgrade my Mac mini from 10.6 to 10.7 to
verify fixing the issue upstream in OpenPGM.

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/96c40fd0/attachment.htm 

From Andrzej.Dworak at cern.ch  Tue Jul 17 23:18:04 2012
From: Andrzej.Dworak at cern.ch (Andrzej Dworak)
Date: Tue, 17 Jul 2012 21:18:04 +0000
Subject: [zeromq-dev] what are the best practices to send
	complex	objects?
In-Reply-To: <CAG+SKy6jivviwWPHPmi3PrhZX3aP7JdTxrRwTZSddocXHK7ibw@mail.gmail.com>
References: <5005164F.9030603@mail.ru>
	<94236CF4-F8CF-492A-AD4C-1A3653FC107B@gmail.com>,
	<CAG+SKy6jivviwWPHPmi3PrhZX3aP7JdTxrRwTZSddocXHK7ibw@mail.gmail.com>
Message-ID: <7D80327D0646EC4794FBAE34D053E8667985866D@PLOXCHG03.cern.ch>

Hi,
I did some extensive testing of serialization libraries (tested C++ and Java implementations) more or less 3 months ago. Two products that I liked the most were MessagePack and Google Protocol Buffers. For my case, where I need to have self-describing serialization, I've chosen MessagePack. The requirements were quite high as this is going to be used by one of a core systems at CERN to operate LHC and other accelerators. So far no problems with MP, it's a great product, well, the same as ZeroMQ :)
Cheers,
Andrzej


________________________________
From: zeromq-dev-bounces at lists.zeromq.org [zeromq-dev-bounces at lists.zeromq.org] on behalf of aalok sood [aalok at ideadevice.com]
Sent: 17 July 2012 12:01
To: ZeroMQ development list
Subject: Re: [zeromq-dev] what are the best practices to send complex objects?

If you are **really** concerned about performance, I suggest you have your own protocol that both sides understand.
This way you send data as it is, the remote side would know how to interpret it.



On Tue, Jul 17, 2012 at 3:27 PM, Cem Karan <cfkaran2 at gmail.com<mailto:cfkaran2 at gmail.com>> wrote:
I've been playing with MessagePack (http://msgpack.org/) for serialization and BLOSC (http://blosc.pytables.org/trac) for compression.  They seem to work pretty well together.  How fast is fast?  BLOSC is designed to move data around in memory faster than memcpy() is able to do it, and I haven't had any complaints about how fast MessagePack works (OTOH, I'm I/O bound, so I tend not to hit any limits).

Thanks,
Cem Karan

On Jul 17, 2012, at 3:37 AM, anti_tenzor wrote:

> Hello, all,
>
> I'd like to use ZMQ to send complex objects between applications (using
> CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
> I am forced to implement some serialization-deserialization mechanism.
>
> But the hole idea is to be as fast as possible! So it looks like this
> algo should be:
> 1. Extremely fast
> 2. Produce very compact byte arrays
>
> Could you please suggest some best practices or libraries (.net), that
> meets these requirements?
>
> How do you send complex ojects via ZMQ in real-life applications?
>
> Thank you in advance!
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org<mailto:zeromq-dev at lists.zeromq.org>
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org<mailto:zeromq-dev at lists.zeromq.org>
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/f82a9c08/attachment.htm 

From ronaldbelliott at gmail.com  Tue Jul 17 23:23:06 2012
From: ronaldbelliott at gmail.com (Ron Elliott)
Date: Tue, 17 Jul 2012 14:23:06 -0700
Subject: [zeromq-dev] what are the best practices to send complex
	objects?
In-Reply-To: <7D80327D0646EC4794FBAE34D053E8667985866D@PLOXCHG03.cern.ch>
References: <5005164F.9030603@mail.ru>
	<94236CF4-F8CF-492A-AD4C-1A3653FC107B@gmail.com>
	<CAG+SKy6jivviwWPHPmi3PrhZX3aP7JdTxrRwTZSddocXHK7ibw@mail.gmail.com>
	<7D80327D0646EC4794FBAE34D053E8667985866D@PLOXCHG03.cern.ch>
Message-ID: <CAMEND1g+RnY2ccypxyr46yNdz1RHkAsQYP+1K5YEWQLab6tOoQ@mail.gmail.com>

I'm not the OP however would it be possible for you to explain why you
chose MessagePack over ProtoBufs? I currently use ProtoBufs myself and
don't have any issues with it however I think this would be interesting to
know
On Jul 17, 2012 2:18 PM, "Andrzej Dworak" <Andrzej.Dworak at cern.ch> wrote:

>  Hi,
> I did some extensive testing of serialization libraries (tested C++ and
> Java implementations) more or less 3 months ago. Two products that I liked
> the most were MessagePack and Google Protocol Buffers. For my case, where I
> need to have self-describing serialization, I've chosen MessagePack. The
> requirements were quite high as this is going to be used by one of a core
> systems at CERN to operate LHC and other accelerators. So far no problems
> with MP, it's a great product, well, the same as ZeroMQ :)
> Cheers,
> Andrzej
>
>
>  ------------------------------
> *From:* zeromq-dev-bounces at lists.zeromq.org [
> zeromq-dev-bounces at lists.zeromq.org] on behalf of aalok sood [
> aalok at ideadevice.com]
> *Sent:* 17 July 2012 12:01
> *To:* ZeroMQ development list
> *Subject:* Re: [zeromq-dev] what are the best practices to send complex
> objects?
>
>  If you are **really** concerned about performance, I suggest you have
> your own protocol that both sides understand.
> This way you send data as it is, the remote side would know how to
> interpret it.
>
>
>
> On Tue, Jul 17, 2012 at 3:27 PM, Cem Karan <cfkaran2 at gmail.com> wrote:
>
>> I've been playing with MessagePack (http://msgpack.org/) for
>> serialization and BLOSC (http://blosc.pytables.org/trac) for
>> compression.  They seem to work pretty well together.  How fast is fast?
>>  BLOSC is designed to move data around in memory faster than memcpy() is
>> able to do it, and I haven't had any complaints about how fast MessagePack
>> works (OTOH, I'm I/O bound, so I tend not to hit any limits).
>>
>> Thanks,
>> Cem Karan
>>
>> On Jul 17, 2012, at 3:37 AM, anti_tenzor wrote:
>>
>> > Hello, all,
>> >
>> > I'd like to use ZMQ to send complex objects between applications (using
>> > CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
>> > I am forced to implement some serialization-deserialization mechanism.
>> >
>> > But the hole idea is to be as fast as possible! So it looks like this
>> > algo should be:
>> > 1. Extremely fast
>> > 2. Produce very compact byte arrays
>> >
>> > Could you please suggest some best practices or libraries (.net), that
>> > meets these requirements?
>> >
>> > How do you send complex ojects via ZMQ in real-life applications?
>> >
>> > Thank you in advance!
>> >
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/0796f455/attachment.htm 

From n1555 at yahoo.fr  Wed Jul 18 00:35:49 2012
From: n1555 at yahoo.fr (Eric)
Date: Tue, 17 Jul 2012 23:35:49 +0100 (BST)
Subject: [zeromq-dev] (no subject)
Message-ID: <1342564549.52542.YahooMailClassic@web29503.mail.ird.yahoo.com>

Trying to use jzmq under android.I generated using zmq3.x procedure, having to change config.sub and config.guess files but generation is OK and match the scheme on?http://www.zeromq.org/build:android.
But I always get unsatisfiedlink error when loading library libjzmq (loadlibrary(jzmq)), whereas i'm able to load libzmq without any problem. I've seen other people asking the same question without answer. Searching for hours over the net, no idea, any help would be appreciated!
Best regards.Eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/2c623e94/attachment.htm 

From victor at iso3103.net  Wed Jul 18 00:48:13 2012
From: victor at iso3103.net (Victor Perron)
Date: Wed, 18 Jul 2012 00:48:13 +0200
Subject: [zeromq-dev] (no subject)
In-Reply-To: <1342564549.52542.YahooMailClassic@web29503.mail.ird.yahoo.com>
References: <1342564549.52542.YahooMailClassic@web29503.mail.ird.yahoo.com>
Message-ID: <CAF9FejH_g-JmM1sdtjJ8HsRs-6WEObNk6ArokVoH67FqzW=H5A@mail.gmail.com>

Same answer as before. I paste it down there.

*Hi,

In that case it seems that you discovered something serious.
I took a look at it, and indeed the assign() symbol from std::string class
seems undefined.
That would mean, Android does _not_ define such a function in its SDK.

Now, the proper way to get around this is a patch to Zeromq 3.x.* *
You'll have to fix those lines (my guess):

./src/ipc_listener.cpp:151:    filename.assign(addr_);
./src/socket_base.cpp:331:    *
*        options.last_endpoint.assign (addr_);
./src/socket_base.cpp:480:    **    options.last_endpoint.assign (addr_);
./src/tcp_address.cpp:503:    **    addr_str.assign (name_, delimiter -
name_);
./src/tcp_address.cpp:504:    **    mask_str.assign (delimiter + 1);
./src/tcp_address.cpp:511:    **    addr_str.assign (name_);

... and try to compile again. If there are no other undefined symbols, that
should work.
That issue did not exist in ZeroMQ 2.2 though, that one didn't make an
extensive use of std::string.

Keep me updated, we'll see for a pull request when I'll be less busy ;)* *

Best regards,*
*

*I'll take a look for an acceptable patch tomorrow, if nobody shows up with
a better suspect/solution of course !
On Wed, Jul 18, 2012 at 12:35 AM, Eric <n1555 at yahoo.fr> wrote:

> Trying to use jzmq under android.
> I generated using zmq3.x procedure, having to change config.sub and
> config.guess files but generation is OK and match the scheme on
> http://www.zeromq.org/build:android.
>
> But I always get unsatisfiedlink error when loading library libjzmq
> (loadlibrary(jzmq)), whereas i'm able to load libzmq without any problem.
> I've seen other people asking the same question without answer. Searching
> for hours over the net, no idea, any help would be appreciated!
>
> Best regards.
> Eric
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/abeaefd2/attachment.htm 

From ronaldbelliott at gmail.com  Wed Jul 18 00:53:52 2012
From: ronaldbelliott at gmail.com (Ron Elliott)
Date: Tue, 17 Jul 2012 15:53:52 -0700
Subject: [zeromq-dev] (no subject)
In-Reply-To: <CAF9FejH_g-JmM1sdtjJ8HsRs-6WEObNk6ArokVoH67FqzW=H5A@mail.gmail.com>
References: <1342564549.52542.YahooMailClassic@web29503.mail.ird.yahoo.com>
	<CAF9FejH_g-JmM1sdtjJ8HsRs-6WEObNk6ArokVoH67FqzW=H5A@mail.gmail.com>
Message-ID: <CAMEND1j7AdYW_Aau6L9+_0PmWP4iWpafH2xnJn6H5vEHTuj6Xw@mail.gmail.com>

perhaps this is an option: http://www.crystax.net/en/android/ndk/7
On Jul 17, 2012 3:48 PM, "Victor Perron" <victor at iso3103.net> wrote:

> Same answer as before. I paste it down there.
>
> *Hi,
>
> In that case it seems that you discovered something serious.
> I took a look at it, and indeed the assign() symbol from std::string class
> seems undefined.
> That would mean, Android does _not_ define such a function in its SDK.
>
> Now, the proper way to get around this is a patch to Zeromq 3.x.* *
> You'll have to fix those lines (my guess):
>
> ./src/ipc_listener.cpp:151:    filename.assign(addr_);
> ./src/socket_base.cpp:331:    *
> *        options.last_endpoint.assign (addr_);
> ./src/socket_base.cpp:480:    **    options.last_endpoint.assign (addr_);
> ./src/tcp_address.cpp:503:    **    addr_str.assign (name_, delimiter -
> name_);
> ./src/tcp_address.cpp:504:    **    mask_str.assign (delimiter + 1);
> ./src/tcp_address.cpp:511:    **    addr_str.assign (name_);
>
> ... and try to compile again. If there are no other undefined symbols,
> that should work.
> That issue did not exist in ZeroMQ 2.2 though, that one didn't make an
> extensive use of std::string.
>
> Keep me updated, we'll see for a pull request when I'll be less busy ;)* *
>
> Best regards,*
> *
>
> *I'll take a look for an acceptable patch tomorrow, if nobody shows up
> with a better suspect/solution of course !
> On Wed, Jul 18, 2012 at 12:35 AM, Eric <n1555 at yahoo.fr> wrote:
>
>> Trying to use jzmq under android.
>> I generated using zmq3.x procedure, having to change config.sub and
>> config.guess files but generation is OK and match the scheme on
>> http://www.zeromq.org/build:android.
>>
>> But I always get unsatisfiedlink error when loading library libjzmq
>> (loadlibrary(jzmq)), whereas i'm able to load libzmq without any problem.
>> I've seen other people asking the same question without answer. Searching
>> for hours over the net, no idea, any help would be appreciated!
>>
>> Best regards.
>> Eric
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>
>
> --
> Victor
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120717/e7e2f972/attachment.htm 

From akeahan at gbtradingllc.com  Wed Jul 18 01:23:32 2012
From: akeahan at gbtradingllc.com (Alex Keahan)
Date: Tue, 17 Jul 2012 19:23:32 -0400
Subject: [zeromq-dev] syncpub/syncsub example losing messages (ZMQ 3.2.0-rc1,
	Solaris 11)
Message-ID: <CACrU9X5XEmbGXnb7bcspCKpfOBdH=dTNir+MPZOt=PRAyTPcBw@mail.gmail.com>

ZMQ 3.2.0-rc1 was built with the latest Sun Studio C/C++ compilers,
with CFLAGS and CXXFLAGS=-fast -library=stlport4.

$ uname -a
SunOS xxxxxx 5.11 11.0 i86pc i386 i86pc
$ CC -V
CC: Sun C++ 5.12 SunOS_i386 2011/11/16
$ cc -V
cc: Sun C 5.12 SunOS_i386 2011/11/16


The syncpub/syncsub example from the guide was adapted to 3.2.0 (see
the code below).

The original version of syncsub would simply get stuck in zmq_recv().
 I added extra code to detect sequence gaps and break out of the loop
whenever there's a discrepancy; now all syncsubs print "received N
updates" where N is somewhat random.   Slowing down the producer fixes
the problem.   Zmq reports no errors at any point.

As a side note, whenever there is a sequence gap, the next sequence
number received by syncsub appears to come from the start of the next
send() buffer sent by syncpub (confirmed by 'truss syncpub')

Any suggestions?

Alex Keahan


Run as follows:

$ ./syncpub &
$ ./syncsub A &
$ ./syncsub B &

You should see something similar to this:

$ ./syncpub &
[1] 6611
$ Waiting for subscribers

$ ./syncsub A &
[2] 6612
$ ./syncsub B &
[3] 6613
$ Broadcasting messages
Message mismatch: received 'Msg #1266' expected 'Msg #1001'
A: received 1000 updates
Message mismatch: received 'Msg #1266' expected 'Msg #1001'
B: received 1000 updates

[2]-  Done                    ./syncsub A
[3]+  Done                    ./syncsub B


syncpub.c:

#include <stdio.h>
#include <strings.h>
#include <unistd.h>
#include <zmq.h>

// Wait for 2 subscribers
#define SUBSCRIBERS_EXPECTED 2

int
main(int argc, char *argv[])
{
  void *context = zmq_ctx_new();
  if (context == 0) {
    fprintf(stderr, "Error: zmq_ctx_new failed: %s\n", zmq_strerror(errno));
    return 1;
  }
  int rc = zmq_ctx_set(context, ZMQ_IO_THREADS, 1);
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_ctx_set failed: %s\n", zmq_strerror(errno));
    return 2;
  }

  // Socket to talk to clients
  void *publisher = zmq_socket(context, ZMQ_PUB);
  if (publisher == 0) {
    fprintf(stderr, "Error: zmq_socket(ZMQ_PUB) failed: %s\n",
zmq_strerror(errno));
    return 4;
  }

  rc = zmq_bind (publisher, "tcp://*:5561");
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_bind failed: %s\n", zmq_strerror(errno));
    return 5;
  }

  // Socket to receive signals
  void *syncservice = zmq_socket (context, ZMQ_REP);
  if (syncservice == 0) {
    fprintf(stderr, "Error: zmq_socket(ZMQ_REP) failed: %s\n",
zmq_strerror(errno));
    return 6;
  }

  rc = zmq_bind (syncservice, "tcp://*:5562");
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_bind failed: %s\n", zmq_strerror(errno));
    return 7;
  }

  // Get synchronization from subscribers
  printf ("Waiting for subscribers\n");
  int subscribers = 0;
  while (subscribers < SUBSCRIBERS_EXPECTED) {
    char tmp[24];

    // - wait for synchronization request
    int size = zmq_recv(syncservice, tmp, sizeof(tmp), 0);
    if (size < 0) {
      fprintf(stderr, "Error: zmq_recv failed: %s\n", zmq_strerror(errno));
      return 8;
    }

    // - send synchronization reply
    rc = zmq_send(syncservice, "", 0, 0);
    if (rc < 0) {
      fprintf(stderr, "Error: zmq_send failed: %s\n", zmq_strerror(errno));
      return 9;
    }

    subscribers++;
  }
  // Now broadcast exactly 100,000 updates followed by END
  printf ("Broadcasting messages\n");
  int update_nbr;
  for (update_nbr = 0; update_nbr < 100000; update_nbr++) {
    char buf[24];
    int len;
    sprintf(buf, "Msg #%d", update_nbr+1);
    len = strlen(buf);

    int size = zmq_send(publisher, buf, len, 0);
    if (size != len) {
      if (errno == EAGAIN) {
        update_nbr--;
        continue;
      }
      fprintf(stderr, "Error: zmq_send failed: %s\n", zmq_strerror(errno));
      return 10;
    }
  }

  int size = zmq_send(publisher, "END", 3, 0);
  if (size != 3) {
    fprintf(stderr, "Error: zmq_send failed: %s\n", zmq_strerror(errno));
    return 11;
  }

  sleep(10);

  zmq_close(publisher);
  zmq_close(syncservice);
  zmq_term (context);

  return 0;
}


syncsub.c:


#include <stdio.h>
#include <strings.h>
#include <unistd.h>
#include <zmq.h>

int
main(int argc, char *argv[])
{
  const char *prefix = (argc > 1 ? argv[1] : "");

  void *context = zmq_ctx_new();
  if (context == 0) {
    fprintf(stderr, "Error: zmq_ctx_new failed: %s\n", zmq_strerror(errno));
    return 1;
  }
  int rc = zmq_ctx_set(context, ZMQ_IO_THREADS, 1);
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_ctx_set failed: %s\n", zmq_strerror(errno));
    return 2;
  }

  // First, connect our subscriber socket
  void *subscriber = zmq_socket(context, ZMQ_SUB);
  if (subscriber == 0) {
    fprintf(stderr, "Error: zmq_socket(ZMQ_SUB) failed: %s\n",
zmq_strerror(errno));
    return 4;
  }

  rc = zmq_connect(subscriber, "tcp://localhost:5561");
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_connect failed: %s\n", zmq_strerror(errno));
    return 5;
  }

  rc = zmq_setsockopt (subscriber, ZMQ_SUBSCRIBE, "", 0);
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_setsockopt failed: %s\n", zmq_strerror(errno));
    return 6;
  }

  // 0MQ is so fast, we need to wait a while??
  sleep (1);

  // Second, synchronize with publisher
  void *syncclient = zmq_socket (context, ZMQ_REQ);
  if (syncclient == 0) {
    fprintf(stderr, "Error: zmq_socket(ZMQ_REQ) failed: %s\n",
zmq_strerror(errno));
    return 7;
  }

  rc = zmq_connect (syncclient, "tcp://localhost:5562");
  if (rc != 0) {
    fprintf(stderr, "Error: zmq_connect failed: %s\n", zmq_strerror(errno));
    return 8;
  }

  // - send a synchronization request
  rc = zmq_send(syncclient, "", 0, 0);
  if (rc < 0) {
    fprintf(stderr, "Error: zmq_send failed: %s\n", zmq_strerror(errno));
    return 9;
  }

  // - wait for synchronization reply
  char buf[24];
  int size = zmq_recv(syncclient, buf, sizeof(buf), 0);
  if (size < 0) {
    fprintf(stderr, "Error: zmq_recv failed: %s\n", zmq_strerror(errno));
    return 10;
  }

  // Third, get our updates and report how many we got
  int update_nbr = 0;
  while (1) {
    size = zmq_recv(subscriber, buf, sizeof(buf), 0);
    if (size < 0) {
      fprintf(stderr, "Error: zmq_recv failed: %s\n", zmq_strerror(errno));
      return 11;
    }

    // Make sure the message numbers match
    char refbuf[24];
    sprintf(refbuf, "Msg #%d", update_nbr+1);

    if (memcmp(buf, refbuf, strlen(refbuf)) != 0) {
      printf("Message mismatch: received '%*.*s' expected '%s'\n",
size, size, buf, refbuf);
      break;
    }

    if (memcmp(buf, "END", 3) == 0) {
      break;
    }
    update_nbr++;
  }
  printf ("%s: received %d updates\n", prefix, update_nbr);

  zmq_close (subscriber);
  zmq_close (syncclient);
  zmq_term (context);

  return 0;
}

From n1555 at yahoo.fr  Wed Jul 18 01:31:19 2012
From: n1555 at yahoo.fr (Eric)
Date: Wed, 18 Jul 2012 00:31:19 +0100 (BST)
Subject: [zeromq-dev] (no subject)
In-Reply-To: <CAMEND1j7AdYW_Aau6L9+_0PmWP4iWpafH2xnJn6H5vEHTuj6Xw@mail.gmail.com>
Message-ID: <1342567879.22044.YahooMailClassic@web29502.mail.ird.yahoo.com>

Thanks,
I replaced assign by clear/append. But it still not works...something make me feel that I don't have any STD lib available :making a call to?arm-linux-androideabi-ld libjzmq.so give me this :
/home/xxx/libjzmq.so: undefined reference to `std::basic_ios<char, std::char_traits<char> >::init(std::basic_streambuf<char, std::char_traits<char> >*)'/home/xxx/libjzmq.so: undefined reference to `__gnu_cxx::__exchange_and_add(int volatile*, int)'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_decrement(std::_Rb_tree_node_base*)'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(std::string const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::compare(char const*) const'/home/xxx/libjzmq.so: undefined reference to `std::locale::locale()'/home/xxx/libjzmq.so: undefined reference to `__aeabi_uldivmod'/home/xxx/libjzmq.so: undefined reference to `std::string::find(char const*, unsigned int, unsigned int) const'/home/xxx/libjzmq.so: undefined reference to `std::string::append(char const*)'/home/xxx/libjzmq.so: undefined reference to
 `std::__throw_bad_alloc()'/home/xxx/libjzmq.so: undefined reference to `std::string::assign(std::string const&)'/home/xxx/libjzmq.so: undefined reference to `vtable for std::basic_stringstream<char, std::char_traits<char>, std::allocator<char> >'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::~basic_string()'/home/xxx/libjzmq.so: undefined reference to `std::basic_stringstream<char, std::char_traits<char>, std::allocator<char> >::~basic_stringstream()'/home/xxx/libjzmq.so: undefined reference to `std::__throw_length_error(char const*)'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_increment(std::_Rb_tree_node_base const*)'/home/xxx/libjzmq.so: undefined reference to `VTT for std::basic_stringstream<char, std::char_traits<char>, std::allocator<char> >'/home/xxx/libjzmq.so: undefined reference to `__gnu_cxx::__atomic_add(int volatile*, int)'/home/xxx/libjzmq.so: undefined
 reference to `std::string::_Rep::_S_empty_rep_storage'/home/xxx/libjzmq.so: undefined reference to `std::ostream& std::ostream::_M_insert<unsigned long>(unsigned long)'/home/xxx/libjzmq.so: undefined reference to `std::ostream::operator<<(int)'/home/xxx/libjzmq.so: undefined reference to `std::ios_base::ios_base()'/home/xxx/libjzmq.so: undefined reference to `std::string::_M_leak()'/home/xxx/libjzmq.so: undefined reference to `std::basic_ostream<char, std::char_traits<char> >& std::__ostream_insert<char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, char const*, int)'/home/xxx/libjzmq.so: undefined reference to `std::string::_M_mutate(unsigned int, unsigned int, unsigned int)'/home/xxx/libjzmq.so: undefined reference to `std::locale::~locale()'/home/xxx/libjzmq.so: undefined reference to `std::string::clear()'/home/xxx/libjzmq.so: undefined reference to `std::__throw_logic_error(char const*)'/home/xxx/libjzmq.so: undefined
 reference to `std::_Rb_tree_insert_and_rebalance(bool, std::_Rb_tree_node_base*, std::_Rb_tree_node_base*, std::_Rb_tree_node_base&)'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(char const*, std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_rebalance_for_erase(std::_Rb_tree_node_base*, std::_Rb_tree_node_base&)'/home/xxx/libjzmq.so: undefined reference to `std::string::_M_leak_hard()'/home/xxx/libjzmq.so: undefined reference to `vtable for __cxxabiv1::__vmi_class_type_info'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(std::string const&, unsigned int, unsigned int)'/home/xxx/libjzmq.so: undefined reference to `vtable for std::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >'/home/xxx/libjzmq.so: undefined reference to `vtable
 for std::basic_streambuf<char, std::char_traits<char> >'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(char const*, unsigned int, std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::ios_base::~ios_base()'/home/xxx/libjzmq.so: undefined reference to `__aeabi_uidivmod'/home/xxx/libjzmq.so: undefined reference to `vtable for std::basic_ios<char, std::char_traits<char> >'/home/xxx/libjzmq.so: undefined reference to `std::basic_istream<char, std::char_traits<char> >::~basic_istream()'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_increment(std::_Rb_tree_node_base*)'/home/xxx/libjzmq.so: undefined reference to `operator delete(void*, std::nothrow_t const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::_Rep::_S_create(unsigned int, unsigned int, std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to
 `std::__throw_out_of_range(char const*)'/home/xxx/libjzmq.so: undefined reference to `operator new(unsigned int, std::nothrow_t const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::_Rep::_M_destroy(std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::append(char const*, unsigned int)'
I'll have a look to what is going on with my configuration.
Best regards.
Eric

--- En date de?: Mer 18.7.12, Ron Elliott <ronaldbelliott at gmail.com> a ?crit?:

De: Ron Elliott <ronaldbelliott at gmail.com>
Objet: Re: [zeromq-dev] (no subject)
?: "ZeroMQ development list" <zeromq-dev at lists.zeromq.org>, victor at iso3103.net
Date: Mercredi 18 juillet 2012, 0h53

perhaps this is an option: http://www.crystax.net/en/android/ndk/7
On Jul 17, 2012 3:48 PM, "Victor Perron" <victor at iso3103.net> wrote:

Same answer as before. I paste it down there.

Hi,

In that case it seems that you discovered something serious. 


I took a look at it, and indeed the assign() symbol from std::string class seems undefined.
That would mean, Android does _not_ define such a function in its SDK.

Now, the proper way to get around this is a patch to Zeromq 3.x.

You'll have to fix those lines (my guess):

./src/ipc_listener.cpp:151:??? filename.assign(addr_);
./src/socket_base.cpp:331:??????????? options.last_endpoint.assign (addr_);



./src/socket_base.cpp:480:??????? options.last_endpoint.assign (addr_);
./src/tcp_address.cpp:503:??????? addr_str.assign (name_, delimiter - name_);
./src/tcp_address.cpp:504:??????? mask_str.assign (delimiter + 1);




./src/tcp_address.cpp:511:??????? addr_str.assign (name_);

... and try to compile again. If there are no other undefined symbols, that should work.
That issue did not exist in ZeroMQ 2.2 though, that one didn't make an extensive use of std::string.




Keep me updated, we'll see for a pull request when I'll be less busy ;)


Best regards,

I'll take a look for an acceptable patch tomorrow, if nobody shows up with a better suspect/solution of course !
On Wed, Jul 18, 2012 at 12:35 AM, Eric <n1555 at yahoo.fr> wrote:



Trying to use jzmq under android.


I generated using zmq3.x procedure, having to change config.sub and config.guess files but generation is OK and match the scheme on?http://www.zeromq.org/build:android.



But I always get unsatisfiedlink error when loading library libjzmq (loadlibrary(jzmq)), whereas i'm able to load libzmq without any problem. I've seen other people asking the same question without answer. Searching for hours over the net, no idea, any help would be appreciated!



Best regards.


Eric


_______________________________________________

zeromq-dev mailing list

zeromq-dev at lists.zeromq.org

http://lists.zeromq.org/mailman/listinfo/zeromq-dev





-- 
Victor



_______________________________________________

zeromq-dev mailing list

zeromq-dev at lists.zeromq.org

http://lists.zeromq.org/mailman/listinfo/zeromq-dev




-----La pi?ce jointe associ?e suit-----

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/b4f6321e/attachment.htm 

From n1555 at yahoo.fr  Wed Jul 18 02:08:23 2012
From: n1555 at yahoo.fr (Eric)
Date: Wed, 18 Jul 2012 01:08:23 +0100 (BST)
Subject: [zeromq-dev] (no subject)
Message-ID: <1342570103.77764.YahooMailClassic@web29504.mail.ird.yahoo.com>

I regenerated with?http://www.crystax.net/en/android/ndk/7, now,?arm-linux-androideabi-ld libjzmq.so?doesn't return any link problem (anymore) with Lib STD : Great!
But still same problem in the android application : loadlibrary works with zmq and still fails (unsatisfies link error) with jzmq.
Don't understand!
Best regards.
--- En date de?: Mer 18.7.12,
 Eric <n1555 at yahoo.fr> a ?crit?:

De: Eric <n1555 at yahoo.fr>
Objet: Re: [zeromq-dev] (no subject)
?: "ZeroMQ development list" <zeromq-dev at lists.zeromq.org>, victor at iso3103.net
Date: Mercredi 18 juillet 2012, 1h31

Thanks,
I replaced assign by clear/append. But it still not works...something make me feel that I don't have any STD lib available :making a call to?arm-linux-androideabi-ld libjzmq.so give me this :
/home/xxx/libjzmq.so: undefined reference to `std::basic_ios<char, std::char_traits<char> >::init(std::basic_streambuf<char,
 std::char_traits<char> >*)'/home/xxx/libjzmq.so: undefined reference to `__gnu_cxx::__exchange_and_add(int volatile*, int)'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_decrement(std::_Rb_tree_node_base*)'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
 >::basic_string(std::string const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::compare(char const*) const'/home/xxx/libjzmq.so: undefined reference to `std::locale::locale()'/home/xxx/libjzmq.so: undefined reference to `__aeabi_uldivmod'/home/xxx/libjzmq.so: undefined reference to `std::string::find(char const*, unsigned int, unsigned int) const'/home/xxx/libjzmq.so: undefined reference to `std::string::append(char const*)'/home/xxx/libjzmq.so: undefined reference to `std::__throw_bad_alloc()'/home/xxx/libjzmq.so: undefined reference to `std::string::assign(std::string const&)'/home/xxx/libjzmq.so: undefined reference to `vtable for std::basic_stringstream<char, std::char_traits<char>, std::allocator<char> >'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>,
 std::allocator<char> >::~basic_string()'/home/xxx/libjzmq.so: undefined reference to `std::basic_stringstream<char, std::char_traits<char>, std::allocator<char> >::~basic_stringstream()'/home/xxx/libjzmq.so: undefined reference to `std::__throw_length_error(char const*)'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_increment(std::_Rb_tree_node_base const*)'/home/xxx/libjzmq.so: undefined reference to `VTT for std::basic_stringstream<char, std::char_traits<char>, std::allocator<char> >'/home/xxx/libjzmq.so: undefined reference to `__gnu_cxx::__atomic_add(int volatile*, int)'/home/xxx/libjzmq.so: undefined reference to `std::string::_Rep::_S_empty_rep_storage'/home/xxx/libjzmq.so: undefined reference to `std::ostream& std::ostream::_M_insert<unsigned long>(unsigned long)'/home/xxx/libjzmq.so:
 undefined reference to `std::ostream::operator<<(int)'/home/xxx/libjzmq.so: undefined reference to `std::ios_base::ios_base()'/home/xxx/libjzmq.so: undefined reference to `std::string::_M_leak()'/home/xxx/libjzmq.so: undefined reference to `std::basic_ostream<char, std::char_traits<char> >& std::__ostream_insert<char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, char const*, int)'/home/xxx/libjzmq.so: undefined reference to `std::string::_M_mutate(unsigned int, unsigned int, unsigned int)'/home/xxx/libjzmq.so: undefined reference to `std::locale::~locale()'/home/xxx/libjzmq.so: undefined reference to `std::string::clear()'/home/xxx/libjzmq.so: undefined reference to `std::__throw_logic_error(char const*)'/home/xxx/libjzmq.so: undefined reference to
 `std::_Rb_tree_insert_and_rebalance(bool, std::_Rb_tree_node_base*, std::_Rb_tree_node_base*, std::_Rb_tree_node_base&)'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(char const*, std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::_Rb_tree_rebalance_for_erase(std::_Rb_tree_node_base*, std::_Rb_tree_node_base&)'/home/xxx/libjzmq.so: undefined reference to `std::string::_M_leak_hard()'/home/xxx/libjzmq.so: undefined reference to `vtable for __cxxabiv1::__vmi_class_type_info'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(std::string const&, unsigned int, unsigned int)'/home/xxx/libjzmq.so: undefined reference to `vtable for
 std::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >'/home/xxx/libjzmq.so: undefined reference to `vtable for std::basic_streambuf<char, std::char_traits<char> >'/home/xxx/libjzmq.so: undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(char const*, unsigned int, std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::ios_base::~ios_base()'/home/xxx/libjzmq.so: undefined reference to `__aeabi_uidivmod'/home/xxx/libjzmq.so: undefined reference to `vtable for std::basic_ios<char, std::char_traits<char> >'/home/xxx/libjzmq.so: undefined reference to `std::basic_istream<char, std::char_traits<char> >::~basic_istream()'/home/xxx/libjzmq.so: undefined reference to
 `std::_Rb_tree_increment(std::_Rb_tree_node_base*)'/home/xxx/libjzmq.so: undefined reference to `operator delete(void*, std::nothrow_t const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::_Rep::_S_create(unsigned int, unsigned int, std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::__throw_out_of_range(char const*)'/home/xxx/libjzmq.so: undefined reference to `operator new(unsigned int, std::nothrow_t const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::_Rep::_M_destroy(std::allocator<char> const&)'/home/xxx/libjzmq.so: undefined reference to `std::string::append(char const*, unsigned int)'
I'll have a look to what is going on with my configuration.
Best regards.
Eric

--- En date de?: Mer
 18.7.12, Ron Elliott <ronaldbelliott at gmail.com> a ?crit?:

De: Ron Elliott <ronaldbelliott at gmail.com>
Objet: Re: [zeromq-dev] (no subject)
?: "ZeroMQ development list" <zeromq-dev at lists.zeromq.org>, victor at iso3103.net
Date: Mercredi 18 juillet 2012, 0h53

perhaps this is an option: http://www.crystax.net/en/android/ndk/7
On Jul 17, 2012 3:48 PM, "Victor Perron" <victor at iso3103.net> wrote:

Same answer as before. I paste it down there.

Hi,

In that case it seems that you discovered something serious. 


I took a look at it, and indeed the assign() symbol from std::string class seems undefined.
That would mean, Android does _not_ define such a function in its SDK.

Now, the proper way to get around this is a patch to Zeromq 3.x.

You'll have to fix those lines (my guess):

./src/ipc_listener.cpp:151:??? filename.assign(addr_);
./src/socket_base.cpp:331:??????????? options.last_endpoint.assign (addr_);



./src/socket_base.cpp:480:??????? options.last_endpoint.assign (addr_);
./src/tcp_address.cpp:503:??????? addr_str.assign (name_, delimiter - name_);
./src/tcp_address.cpp:504:??????? mask_str.assign (delimiter + 1);




./src/tcp_address.cpp:511:??????? addr_str.assign (name_);

... and try to compile again. If there are no other undefined symbols, that should work.
That issue did not exist in ZeroMQ 2.2 though, that one didn't make an extensive use of std::string.




Keep me updated, we'll see for a pull request when I'll be less busy ;)


Best regards,

I'll take a look for an acceptable patch tomorrow, if nobody shows up with a better suspect/solution of course !
On Wed, Jul 18, 2012 at 12:35 AM, Eric <n1555 at yahoo.fr> wrote:



Trying to use jzmq under android.


I generated using zmq3.x procedure, having to change config.sub and config.guess files but generation is OK and match the scheme on?http://www.zeromq.org/build:android.



But I always get unsatisfiedlink error when loading library libjzmq (loadlibrary(jzmq)), whereas i'm able to load libzmq without any problem. I've seen other people asking the same question without answer. Searching for hours over the net, no idea, any help would be appreciated!



Best regards.


Eric


_______________________________________________

zeromq-dev mailing list

zeromq-dev at lists.zeromq.org

http://lists.zeromq.org/mailman/listinfo/zeromq-dev





-- 
Victor



_______________________________________________

zeromq-dev mailing list

zeromq-dev at lists.zeromq.org

http://lists.zeromq.org/mailman/listinfo/zeromq-dev




-----La pi?ce jointe associ?e suit-----

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-----La pi?ce jointe associ?e suit-----

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/9e4a2648/attachment.htm 

From susantharakan83 at gmail.com  Wed Jul 18 09:10:09 2012
From: susantharakan83 at gmail.com (Susan Tharakan)
Date: Wed, 18 Jul 2012 12:40:09 +0530
Subject: [zeromq-dev] Zeromq 3.2 pub sub in a large production environment
Message-ID: <CAMjz1dT+HP8M0TqkZ7p1cjiatRFbpB0nq137jvhnGtb_Z2bvUQ@mail.gmail.com>

Hi,

I am planning to use zeromq 3.2 pub sub in our production systems and
wanted to check with the community. How stable is zeromq 3.2 for production
use? Has anybody used it in production systems for pub sub model.

For our application the following two requirements are critical

1. Minimize latency
2. Avoid data loss


Currently I'm using a slightly modified version of zeromq 2.0.8 in
production for a 400 machine cluster with reasonable success. The major
issue we've had with zeromq 2.x are

1. High bandwidth utilization during high traffic scenarios because of the
lack of subscriber filtering, possibly resulting in buffer over runs at
producer and drops[data loss].


2. Loosing messages when the subscribers cant keep pace with the producers.
Using the SWAP option in zeromq 2.0 has saved the day in some specific
scenarios as it provided some cushion with  spurty traffic.

With zeromq 3.x the SWAP option seems to be discontinued, which is a big
concern for our usecase, but I'm guessing the buffered queue lengths will
also go down as only required msgs are kept in queue now. Any work arounds
for this problem?

The other important requirement we have is to be able to monitor what's
happening inside of zeromq. We'd done it by writing one of our custom apis
for 2.0 which provided some insight into how much data was buffered at each
end. How easy is it to do this for zeromq 3.x / any plans for it in the zmq
roadmap?

Thanks for your time
Susan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/d805d621/attachment.htm 

From mimir at newmail.ru  Wed Jul 18 09:49:08 2012
From: mimir at newmail.ru (moteus)
Date: Wed, 18 Jul 2012 07:49:08 +0000 (UTC)
Subject: [zeromq-dev] Socket use from several threads
References: <loom.20120717T130445-42@post.gmane.org>
	<75FBA2FE-975A-4949-A1A3-86251B2E8AB2@chuckremes.com>
	<loom.20120717T160608-968@post.gmane.org>
	<22DB4BFF-7566-4C54-B0A3-E7F0DDC54AA8@chuckremes.com>
Message-ID: <loom.20120718T093650-292@post.gmane.org>

Chuck Remes <lists <at> chuckremes.com> writes:


> Oh, so the threads are short-lived then. In that case, you should create a
pool of them and have each thread
> "check out" the socket from the pool when it needs to use it and check it back
in again when the thread exits.
> If the pool is protected by a mechanism (like a mutex) that employs a full
memory barrier, then it will be safe.

Thanks.
And one more question. 
Can I start recv in one thread but get result in other?
For example. I use DEALER socket instead REQ and get timeout on recive.
I put socket back to pool. Another thread get this socket from pool and send 
it's own request. This thread can recognize its response and ignore other
responses. 
Or better just to recreate socket like described in guide?




From victor at iso3103.net  Wed Jul 18 09:55:42 2012
From: victor at iso3103.net (Victor Perron)
Date: Wed, 18 Jul 2012 09:55:42 +0200
Subject: [zeromq-dev] (no subject)
In-Reply-To: <1342567879.22044.YahooMailClassic@web29502.mail.ird.yahoo.com>
References: <CAMEND1j7AdYW_Aau6L9+_0PmWP4iWpafH2xnJn6H5vEHTuj6Xw@mail.gmail.com>
	<1342567879.22044.YahooMailClassic@web29502.mail.ird.yahoo.com>
Message-ID: <CAF9FejHDjrofy=7w=iYGdqyMdcpfH1NuhbjugrA6ePC54wd3ZQ@mail.gmail.com>

OK.

The libstdc++ shared object from the NDK defines very few symbols, around
twenty, mostly for what google claims to support: exceptions and such.
In order to use the STL, one must use their "stlport" beta library which I
believe is included since NDKr5.
The shared object is present on my SGSIII, but I think that maybe it's
better to statically include the needed symbols in libzmq to prevent future
issues (and as long as it's in beta and that older devices mat use it)

The other way to go would be to remove all those STL calls from libzmq (as
2.2 was doing) but... ;)

Thing is, I'm afraid that those functions you reported not found are not
even in Android's stlport. And I'm not happy yo use another, third-party C
library.

I'll update the wiki about this issue, and I think you should stick to
zeromq2.2 before that; that one has been properly tested as working on
Android.

Regards,

On Wed, Jul 18, 2012 at 1:31 AM, Eric <n1555 at yahoo.fr> wrote:

> Thanks,
>
> I replaced assign by clear/append. But it still not works...something make
> me feel that I don't have any STD lib available :
> making a call to arm-linux-androideabi-ld libjzmq.so give me this :
>
> /home/xxx/libjzmq.so: undefined reference to `std::basic_ios<char,
> std::char_traits<char> >::init(std::basic_streambuf<char,
> std::char_traits<char> >*)'
> /home/xxx/libjzmq.so: undefined reference to
> `__gnu_cxx::__exchange_and_add(int volatile*, int)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::_Rb_tree_decrement(std::_Rb_tree_node_base*)'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_string<char,
> std::char_traits<char>, std::allocator<char> >::basic_string(std::string
> const&)'
> /home/xxx/libjzmq.so: undefined reference to `std::string::compare(char
> const*) const'
> /home/xxx/libjzmq.so: undefined reference to `std::locale::locale()'
> /home/xxx/libjzmq.so: undefined reference to `__aeabi_uldivmod'
> /home/xxx/libjzmq.so: undefined reference to `std::string::find(char
> const*, unsigned int, unsigned int) const'
> /home/xxx/libjzmq.so: undefined reference to `std::string::append(char
> const*)'
> /home/xxx/libjzmq.so: undefined reference to `std::__throw_bad_alloc()'
> /home/xxx/libjzmq.so: undefined reference to
> `std::string::assign(std::string const&)'
> /home/xxx/libjzmq.so: undefined reference to `vtable for
> std::basic_stringstream<char, std::char_traits<char>, std::allocator<char>
> >'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_string<char,
> std::char_traits<char>, std::allocator<char> >::~basic_string()'
> /home/xxx/libjzmq.so: undefined reference to
> `std::basic_stringstream<char, std::char_traits<char>, std::allocator<char>
> >::~basic_stringstream()'
> /home/xxx/libjzmq.so: undefined reference to
> `std::__throw_length_error(char const*)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::_Rb_tree_increment(std::_Rb_tree_node_base const*)'
> /home/xxx/libjzmq.so: undefined reference to `VTT for
> std::basic_stringstream<char, std::char_traits<char>, std::allocator<char>
> >'
> /home/xxx/libjzmq.so: undefined reference to `__gnu_cxx::__atomic_add(int
> volatile*, int)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::string::_Rep::_S_empty_rep_storage'
> /home/xxx/libjzmq.so: undefined reference to `std::ostream&
> std::ostream::_M_insert<unsigned long>(unsigned long)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::ostream::operator<<(int)'
> /home/xxx/libjzmq.so: undefined reference to `std::ios_base::ios_base()'
> /home/xxx/libjzmq.so: undefined reference to `std::string::_M_leak()'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_ostream<char,
> std::char_traits<char> >& std::__ostream_insert<char,
> std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char>
> >&, char const*, int)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::string::_M_mutate(unsigned int, unsigned int, unsigned int)'
> /home/xxx/libjzmq.so: undefined reference to `std::locale::~locale()'
> /home/xxx/libjzmq.so: undefined reference to `std::string::clear()'
> /home/xxx/libjzmq.so: undefined reference to
> `std::__throw_logic_error(char const*)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::_Rb_tree_insert_and_rebalance(bool, std::_Rb_tree_node_base*,
> std::_Rb_tree_node_base*, std::_Rb_tree_node_base&)'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_string<char,
> std::char_traits<char>, std::allocator<char> >::basic_string(char const*,
> std::allocator<char> const&)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::_Rb_tree_rebalance_for_erase(std::_Rb_tree_node_base*,
> std::_Rb_tree_node_base&)'
> /home/xxx/libjzmq.so: undefined reference to `std::string::_M_leak_hard()'
> /home/xxx/libjzmq.so: undefined reference to `vtable for
> __cxxabiv1::__vmi_class_type_info'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_string<char,
> std::char_traits<char>, std::allocator<char> >::basic_string(std::string
> const&, unsigned int, unsigned int)'
> /home/xxx/libjzmq.so: undefined reference to `vtable for
> std::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >'
> /home/xxx/libjzmq.so: undefined reference to `vtable for
> std::basic_streambuf<char, std::char_traits<char> >'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_string<char,
> std::char_traits<char>, std::allocator<char> >::basic_string(char const*,
> unsigned int, std::allocator<char> const&)'
> /home/xxx/libjzmq.so: undefined reference to `std::ios_base::~ios_base()'
> /home/xxx/libjzmq.so: undefined reference to `__aeabi_uidivmod'
> /home/xxx/libjzmq.so: undefined reference to `vtable for
> std::basic_ios<char, std::char_traits<char> >'
> /home/xxx/libjzmq.so: undefined reference to `std::basic_istream<char,
> std::char_traits<char> >::~basic_istream()'
> /home/xxx/libjzmq.so: undefined reference to
> `std::_Rb_tree_increment(std::_Rb_tree_node_base*)'
> /home/xxx/libjzmq.so: undefined reference to `operator delete(void*,
> std::nothrow_t const&)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::string::_Rep::_S_create(unsigned int, unsigned int,
> std::allocator<char> const&)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::__throw_out_of_range(char const*)'
> /home/xxx/libjzmq.so: undefined reference to `operator new(unsigned int,
> std::nothrow_t const&)'
> /home/xxx/libjzmq.so: undefined reference to
> `std::string::_Rep::_M_destroy(std::allocator<char> const&)'
> /home/xxx/libjzmq.so: undefined reference to `std::string::append(char
> const*, unsigned int)'
>
> I'll have a look to what is going on with my configuration.
>
> Best regards.
>
> Eric
>
>
> --- En date de : *Mer 18.7.12, Ron Elliott <ronaldbelliott at gmail.com>* a
> ?crit :
>
>
> De: Ron Elliott <ronaldbelliott at gmail.com>
> Objet: Re: [zeromq-dev] (no subject)
> ?: "ZeroMQ development list" <zeromq-dev at lists.zeromq.org>,
> victor at iso3103.net
> Date: Mercredi 18 juillet 2012, 0h53
>
>
> perhaps this is an option: http://www.crystax.net/en/android/ndk/7
> On Jul 17, 2012 3:48 PM, "Victor Perron" <victor at iso3103.net<http://mc/compose?to=victor at iso3103.net>>
> wrote:
>
> Same answer as before. I paste it down there.
>
> *Hi,
>
> In that case it seems that you discovered something serious.
> I took a look at it, and indeed the assign() symbol from std::string class
> seems undefined.
> That would mean, Android does _not_ define such a function in its SDK.
>
> Now, the proper way to get around this is a patch to Zeromq 3.x.* *
> You'll have to fix those lines (my guess):
>
> ./src/ipc_listener.cpp:151:    filename.assign(addr_);
> ./src/socket_base.cpp:331:    *
> *        options.last_endpoint.assign (addr_);
> ./src/socket_base.cpp:480:    **    options.last_endpoint.assign (addr_);
> ./src/tcp_address.cpp:503:    **    addr_str.assign (name_, delimiter -
> name_);
> ./src/tcp_address.cpp:504:    **    mask_str.assign (delimiter + 1);
> ./src/tcp_address.cpp:511:    **    addr_str.assign (name_);
>
> ... and try to compile again. If there are no other undefined symbols,
> that should work.
> That issue did not exist in ZeroMQ 2.2 though, that one didn't make an
> extensive use of std::string.
>
> Keep me updated, we'll see for a pull request when I'll be less busy ;)* *
>
> Best regards,*
> *
>
> *I'll take a look for an acceptable patch tomorrow, if nobody shows up
> with a better suspect/solution of course !
> On Wed, Jul 18, 2012 at 12:35 AM, Eric <n1555 at yahoo.fr<http://mc/compose?to=n1555 at yahoo.fr>
> > wrote:
>
> Trying to use jzmq under android.
> I generated using zmq3.x procedure, having to change config.sub and
> config.guess files but generation is OK and match the scheme on
> http://www.zeromq.org/build:android.
>
> But I always get unsatisfiedlink error when loading library libjzmq
> (loadlibrary(jzmq)), whereas i'm able to load libzmq without any problem.
> I've seen other people asking the same question without answer. Searching
> for hours over the net, no idea, any help would be appreciated!
>
> Best regards.
> Eric
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org<http://mc/compose?to=zeromq-dev at lists.zeromq.org>
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
>
>
> --
> Victor
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org<http://mc/compose?to=zeromq-dev at lists.zeromq.org>
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
> -----La pi?ce jointe associ?e suit-----
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org<http://mc/compose?to=zeromq-dev at lists.zeromq.org>
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/b271ab87/attachment.htm 

From lists at chuckremes.com  Wed Jul 18 13:33:58 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Wed, 18 Jul 2012 06:33:58 -0500
Subject: [zeromq-dev] Socket use from several threads
In-Reply-To: <loom.20120718T093650-292@post.gmane.org>
References: <loom.20120717T130445-42@post.gmane.org>
	<75FBA2FE-975A-4949-A1A3-86251B2E8AB2@chuckremes.com>
	<loom.20120717T160608-968@post.gmane.org>
	<22DB4BFF-7566-4C54-B0A3-E7F0DDC54AA8@chuckremes.com>
	<loom.20120718T093650-292@post.gmane.org>
Message-ID: <78BD0133-7EB5-4ACD-A6C4-CDAD1A4CB0B9@chuckremes.com>

On Jul 18, 2012, at 2:49 AM, moteus wrote:

> Chuck Remes <lists <at> chuckremes.com> writes:
> 
> 
>> Oh, so the threads are short-lived then. In that case, you should create a
> pool of them and have each thread
>> "check out" the socket from the pool when it needs to use it and check it back
> in again when the thread exits.
>> If the pool is protected by a mechanism (like a mutex) that employs a full
> memory barrier, then it will be safe.
> 
> Thanks.
> And one more question. 
> Can I start recv in one thread but get result in other?
> For example. I use DEALER socket instead REQ and get timeout on recive.
> I put socket back to pool. Another thread get this socket from pool and send 
> it's own request. This thread can recognize its response and ignore other
> responses. 
> Or better just to recreate socket like described in guide?

I don't know. Try it out and let us know how it works.

cr

From Sharon.Ben-Asher at avg.com  Wed Jul 18 14:53:34 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Wed, 18 Jul 2012 12:53:34 +0000
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D64F1@ex10czdmb01>

Hi,

Well, after realizing its an issue with STL, I spent half the day on the net researching Android NDK and C++ STL.
Indeed, the default implementation is "minimal" according to the docs.
The issue is really an issue since we are using the Android NDK as standalone compiler.  As a full build env, you can specify in the Application.mk (their version of makefile) that you want to use STL.  They even support several versions of the library. 
So perhaps someone needs to repackage the whole directory tree into an Android project.  I think I lack the knowledge of how to do it. 

However, I did find that the standalone toolchain does come with gunstl_shared lib.  And I even managed to specify the proper argument to configure and to build new shared object files. 
However, now when I try to run the APK, I get an error about not finding libgnustl_shared.so.  tried to put it together with libjzmq.so, but to no avail.
I am running out of time with this experiment, they want to move to an all-java solution. So if time permits, I will try to build the 2.2 version.

Sharon.

-----Original Message-----
From: Victor Perron <victor at iso3103.net>
Subject: Re: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 14
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org>
Message-ID:
	<CAF9FejHK-9LOrwhHMrinkeaiq5AkwAcxijVV9K6X45j5d__EaQ at mail.gmail.com>
Content-Type: text/plain; charset="iso-8859-1"

Hi,

In that case it seems that you discovered something serious.
I took a look at it, and indeed the assign() symbol from std::string class seems undefined.
That would mean, Android does _not_ define such a function in its SDK.

Now, the proper way to get around this is a patch to Zeromq 3.x.
You'll have to fix those lines (my guess):

./src/ipc_listener.cpp:151:    filename.assign(addr_);
./src/socket_base.cpp:331:            options.last_endpoint.assign (addr_);
./src/socket_base.cpp:480:        options.last_endpoint.assign (addr_);
./src/tcp_address.cpp:503:        addr_str.assign (name_, delimiter -
name_);
./src/tcp_address.cpp:504:        mask_str.assign (delimiter + 1);
./src/tcp_address.cpp:511:        addr_str.assign (name_);

... and try to compile again. If there are no other undefined symbols, that should work.
That issue did not exist in ZeroMQ 2.2 though, that one didn't make an extensive use of std::string.

Keep me updated, we'll see for a pull request when I'll be less busy ;)

Best regards,


From Sharon.Ben-Asher at avg.com  Wed Jul 18 14:55:00 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Wed, 18 Jul 2012 12:55:00 +0000
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D6502@ex10czdmb01>

Forget to mention that I didn't know what you meant by "fix those lines"
How do I fix them?

-----Original Message-----
From: Victor Perron <victor at iso3103.net>
Subject: Re: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 14
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org>
Message-ID:
	<CAF9FejHK-9LOrwhHMrinkeaiq5AkwAcxijVV9K6X45j5d__EaQ at mail.gmail.com>
Content-Type: text/plain; charset="iso-8859-1"

Hi,

In that case it seems that you discovered something serious.
I took a look at it, and indeed the assign() symbol from std::string class seems undefined.
That would mean, Android does _not_ define such a function in its SDK.

Now, the proper way to get around this is a patch to Zeromq 3.x.
You'll have to fix those lines (my guess):

./src/ipc_listener.cpp:151:    filename.assign(addr_);
./src/socket_base.cpp:331:            options.last_endpoint.assign (addr_);
./src/socket_base.cpp:480:        options.last_endpoint.assign (addr_);
./src/tcp_address.cpp:503:        addr_str.assign (name_, delimiter -
name_);
./src/tcp_address.cpp:504:        mask_str.assign (delimiter + 1);
./src/tcp_address.cpp:511:        addr_str.assign (name_);

... and try to compile again. If there are no other undefined symbols, that should work.
That issue did not exist in ZeroMQ 2.2 though, that one didn't make an extensive use of std::string.

Keep me updated, we'll see for a pull request when I'll be less busy ;)

Best regards,


From victor at iso3103.net  Wed Jul 18 15:12:43 2012
From: victor at iso3103.net (Victor Perron)
Date: Wed, 18 Jul 2012 15:12:43 +0200
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B384D6502@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B384D6502@ex10czdmb01>
Message-ID: <CAF9FejH05+4X0p6HLoRgsVR-FtshPjAQGt7L0kDkW9BtLn37mQ@mail.gmail.com>

I thought, not use 'assign'. But the issue has been reported and discussed
in the Zeromq dev list.
I changed the Wiki today; you should stick to 2.2 for now, because a proper
fix is out of quick reach with 3.x .

It's either a lot of changes in the instructions in order to make zeromq
use stlport library, or get rid of the STL functions almost entirely.
Even using the STL has very little chances to succeed and will need heavy
patches to zeromq, would it only be for the headers inclusion; but I'm also
unsure that everything zeromq uses is defined by stlport.
Moreover, stlport is beta, so it's quite dangerous to rely on it.

I'd appreciate if somebody really involved into libzmq 3.x core could
comment about this...

On Wed, Jul 18, 2012 at 2:55 PM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com>wrote:

> Forget to mention that I didn't know what you meant by "fix those lines"
> How do I fix them?
>
> -----Original Message-----
> From: Victor Perron <victor at iso3103.net>
> Subject: Re: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 14
> To: ZeroMQ development list <zeromq-dev at lists.zeromq.org>
> Message-ID:
>         <
> CAF9FejHK-9LOrwhHMrinkeaiq5AkwAcxijVV9K6X45j5d__EaQ at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi,
>
> In that case it seems that you discovered something serious.
> I took a look at it, and indeed the assign() symbol from std::string class
> seems undefined.
> That would mean, Android does _not_ define such a function in its SDK.
>
> Now, the proper way to get around this is a patch to Zeromq 3.x.
> You'll have to fix those lines (my guess):
>
> ./src/ipc_listener.cpp:151:    filename.assign(addr_);
> ./src/socket_base.cpp:331:            options.last_endpoint.assign (addr_);
> ./src/socket_base.cpp:480:        options.last_endpoint.assign (addr_);
> ./src/tcp_address.cpp:503:        addr_str.assign (name_, delimiter -
> name_);
> ./src/tcp_address.cpp:504:        mask_str.assign (delimiter + 1);
> ./src/tcp_address.cpp:511:        addr_str.assign (name_);
>
> ... and try to compile again. If there are no other undefined symbols,
> that should work.
> That issue did not exist in ZeroMQ 2.2 though, that one didn't make an
> extensive use of std::string.
>
> Keep me updated, we'll see for a pull request when I'll be less busy ;)
>
> Best regards,
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/c89d1cf3/attachment.htm 

From Andrzej.Dworak at cern.ch  Wed Jul 18 16:12:20 2012
From: Andrzej.Dworak at cern.ch (Andrzej Dworak)
Date: Wed, 18 Jul 2012 14:12:20 +0000
Subject: [zeromq-dev] what are the best practices to send
	complex	objects?
In-Reply-To: <CAMEND1g+RnY2ccypxyr46yNdz1RHkAsQYP+1K5YEWQLab6tOoQ@mail.gmail.com>
References: <5005164F.9030603@mail.ru>
	<94236CF4-F8CF-492A-AD4C-1A3653FC107B@gmail.com>
	<CAG+SKy6jivviwWPHPmi3PrhZX3aP7JdTxrRwTZSddocXHK7ibw@mail.gmail.com>
	<7D80327D0646EC4794FBAE34D053E8667985866D@PLOXCHG03.cern.ch>
	<CAMEND1g+RnY2ccypxyr46yNdz1RHkAsQYP+1K5YEWQLab6tOoQ@mail.gmail.com>
Message-ID: <7D80327D0646EC4794FBAE34D053E866798587C9@PLOXCHG03.cern.ch>

Hi Ron,

Sure, but, as I said before - I liked both, and both are great. I liked them far more than anything else. So, I don't claim one is better than another. What, in my case, made me chose MP over PB in the end is:

1.       I need to send bool, int8,16,32,64, float, double, string, and arrays of these (various lengths from 10 up to even a few millions)

2.       I need to encode types of the data sent in each message.

3.       I've prototyped simple tests for ser/deser in the way it will be used using both libs

4.       Running the tests on supported platforms, on average, MP was 2x faster than PB

Probably, in a normal situation (without extra encoding of the types) PB would do closer to MP.

MP is pretty stable and it gains in popularity but AFAIK there are no big players behind. PB is supported and used by Google, so not much worries about stability there :)

Cheers,
Andrzej


From: zeromq-dev-bounces at lists.zeromq.org [mailto:zeromq-dev-bounces at lists.zeromq.org] On Behalf Of Ron Elliott
Sent: 17 July 2012 23:23
To: ZeroMQ development list
Subject: Re: [zeromq-dev] what are the best practices to send complex objects?


I'm not the OP however would it be possible for you to explain why you chose MessagePack over ProtoBufs? I currently use ProtoBufs myself and don't have any issues with it however I think this would be interesting to know
On Jul 17, 2012 2:18 PM, "Andrzej Dworak" <Andrzej.Dworak at cern.ch<mailto:Andrzej.Dworak at cern.ch>> wrote:
Hi,
I did some extensive testing of serialization libraries (tested C++ and Java implementations) more or less 3 months ago. Two products that I liked the most were MessagePack and Google Protocol Buffers. For my case, where I need to have self-describing serialization, I've chosen MessagePack. The requirements were quite high as this is going to be used by one of a core systems at CERN to operate LHC and other accelerators. So far no problems with MP, it's a great product, well, the same as ZeroMQ :)
Cheers,
Andrzej


________________________________
From: zeromq-dev-bounces at lists.zeromq.org<mailto:zeromq-dev-bounces at lists.zeromq.org> [zeromq-dev-bounces at lists.zeromq.org<mailto:zeromq-dev-bounces at lists.zeromq.org>] on behalf of aalok sood [aalok at ideadevice.com<mailto:aalok at ideadevice.com>]
Sent: 17 July 2012 12:01
To: ZeroMQ development list
Subject: Re: [zeromq-dev] what are the best practices to send complex objects?
If you are **really** concerned about performance, I suggest you have your own protocol that both sides understand.
This way you send data as it is, the remote side would know how to interpret it.


On Tue, Jul 17, 2012 at 3:27 PM, Cem Karan <cfkaran2 at gmail.com<mailto:cfkaran2 at gmail.com>> wrote:
I've been playing with MessagePack (http://msgpack.org/) for serialization and BLOSC (http://blosc.pytables.org/trac) for compression.  They seem to work pretty well together.  How fast is fast?  BLOSC is designed to move data around in memory faster than memcpy() is able to do it, and I haven't had any complaints about how fast MessagePack works (OTOH, I'm I/O bound, so I tend not to hit any limits).

Thanks,
Cem Karan

On Jul 17, 2012, at 3:37 AM, anti_tenzor wrote:

> Hello, all,
>
> I'd like to use ZMQ to send complex objects between applications (using
> CLRZMQ c# wrapper). Since it is possible to send byte[] arrays only,
> I am forced to implement some serialization-deserialization mechanism.
>
> But the hole idea is to be as fast as possible! So it looks like this
> algo should be:
> 1. Extremely fast
> 2. Produce very compact byte arrays
>
> Could you please suggest some best practices or libraries (.net), that
> meets these requirements?
>
> How do you send complex ojects via ZMQ in real-life applications?
>
> Thank you in advance!
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org<mailto:zeromq-dev at lists.zeromq.org>
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org<mailto:zeromq-dev at lists.zeromq.org>
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org<mailto:zeromq-dev at lists.zeromq.org>
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/e3cb5141/attachment.htm 

From ian.barber at gmail.com  Wed Jul 18 21:00:12 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Wed, 18 Jul 2012 20:00:12 +0100
Subject: [zeromq-dev] A query about implementing aborts in request reply
	broker
In-Reply-To: <CAG+SKy6ybC3esb=ob-Evefxj1ZS0AkCQssk8sdmEO8scq3NE1Q@mail.gmail.com>
References: <CAG+SKy5t68xpGrgFdFoWoz9SjDZv9XjuK8A0-TX8OYTGStbdtg@mail.gmail.com>
	<CAEYnEwE6-VwvMyikB+PFeTB9kS_mNf5qLgubKL0iy6eq0Q+05Q@mail.gmail.com>
	<CAG+SKy6ybC3esb=ob-Evefxj1ZS0AkCQssk8sdmEO8scq3NE1Q@mail.gmail.com>
Message-ID: <CAEYnEwEuWD8kmcYfmrPfQ_G1_KOKDR0c8c-9c6FOL0ejspKQkg@mail.gmail.com>

On Mon, Jul 16, 2012 at 11:17 AM, aalok sood <aalok at ideadevice.com> wrote:

> Thanks for your suggestion Ian.
>
> I have a similar model in place, but the issue is how do I find out how
> many responses are going to come in on my pull socket.
>
> In essence, can I somehow find out how many backend workers are connected
> to my broker.
>
>
There is no easy way to do that - the idea with pub sub is that you
generally want to send a message to everyone and no worry precisely about
who the subscribers are. In your case, it might be better to have a REQ-REP
model - clients make a REQ when they're ready to do some work, your server
has a ROUTER and you track the ids, and you send out a message to each one
when you have something to do, then you can track the state internally. If
you wanted to keep the pub/pull model, you could have a worker send a
message to the pull saying that they are starting some work, and use that
to track the number you expect. Of course, you have to consider the
possibility of worker failure - so if a worker starts a job then dies, what
is you action going to be around that, both in the regular and cancellation
cases.

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/78d690c1/attachment.htm 

From bh at udev.org  Wed Jul 18 22:44:57 2012
From: bh at udev.org (Benjamin Henrion)
Date: Wed, 18 Jul 2012 22:44:57 +0200
Subject: [zeromq-dev] 0mq at TheStangeLoop, 23-25 Sept 2012
Message-ID: <CANjd3nfCUqSGov8te1zKO7LGmtBhG-6qztDiJDtriyUzrfwf+Q@mail.gmail.com>

0mq at TheStangeLoop, 23-25 Sept 2012:

https://thestrangeloop.com/sessions/software-architecture-using-zeromq

-- 
Benjamin Henrion <bhenrion at ffii.org>
FFII Brussels - +32-484-566109 - +32-2-3500762
"In July 2005, after several failed attempts to legalise software
patents in Europe, the patent establishment changed its strategy.
Instead of explicitly seeking to sanction the patentability of
software, they are now seeking to create a central European patent
court, which would establish and enforce patentability rules in their
favor, without any possibility of correction by competing courts or
democratically elected legislators."

From akeahan at gbtradingllc.com  Wed Jul 18 22:49:32 2012
From: akeahan at gbtradingllc.com (Alex Keahan)
Date: Wed, 18 Jul 2012 20:49:32 +0000 (UTC)
Subject: [zeromq-dev]
	=?utf-8?q?syncpub/syncsub_example_losing_messages_?=
	=?utf-8?b?KFpNUSAzLjIuMC1yYzEsCVNvbGFyaXMgMTEp?=
References: <CACrU9X5XEmbGXnb7bcspCKpfOBdH=dTNir+MPZOt=PRAyTPcBw@mail.gmail.com>
Message-ID: <loom.20120718T224314-618@post.gmane.org>

Alex Keahan <akeahan <at> gbtradingllc.com> writes:

> $ ./syncpub &
> [1] 6611
> $ Waiting for subscribers
> 
> $ ./syncsub A &
> [2] 6612
> $ ./syncsub B &
> [3] 6613
> $ Broadcasting messages
> Message mismatch: received 'Msg #1266' expected 'Msg #1001'
> A: received 1000 updates
> Message mismatch: received 'Msg #1266' expected 'Msg #1001'
> B: received 1000 updates
> 


Whoops.   The high water mark for outbound messages was changed from 0 (no
limit) in 2.2 to 1000 in 3.2.

Not a bug, please disregard.



From Sharon.Ben-Asher at avg.com  Thu Jul 19 08:53:17 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Thu, 19 Jul 2012 06:53:17 +0000
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D6585@ex10czdmb01>

Hi,

Today, I tried to work with v2.2.0:
followed the instructions as close as I can, the build was successful, as far as I could see.
However, when running in the device, encountered this problem:
Cannot load library: reloc_library[1289]:   157 cannot locate '_ZNSt9basic_iosIcSt11char_traitsIcEE4initEPSt15basic_streambufIcS1_E'...

Thanks,
Sharon.

From bluerocketdelivery at gmail.com  Wed Jul 18 23:31:29 2012
From: bluerocketdelivery at gmail.com (Ron Elliott)
Date: Wed, 18 Jul 2012 14:31:29 -0700
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
	<CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
Message-ID: <513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>

Does version 3.x use the same version of openpgm that compiles on mac? Looks like it's now just easier to upgrade if this issue doesn't exist in v3.

Thanks for your response

On Jul 17, 2012, at 12:27 PM, Steven McCoy wrote:

> On 17 July 2012 02:22, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
> Using the downloaded stable version 2.2 code I cannot get ZMQ to compile with the --with-pgm option. The project is too far along to make the switch to v3 so that's not an option atm. Any other solutions or is ZMQ dead for me?
> 
> 
> It's a known issue, I have to upgrade my Mac mini from 10.6 to 10.7 to verify fixing the issue upstream in OpenPGM.
> 
> -- 
> Steve-o 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120718/d25ef56c/attachment.htm 

From victor at iso3103.net  Thu Jul 19 10:03:55 2012
From: victor at iso3103.net (Victor Perron)
Date: Thu, 19 Jul 2012 10:03:55 +0200
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B384D6585@ex10czdmb01>
References: <671E27F33F2F25499F78A4AFA494E63B384D6585@ex10czdmb01>
Message-ID: <CAF9FejHxad-UXFthBfZcRtU-MY_c9883cymGvGvrOnH8_ccq=w@mail.gmail.com>

This is awkward. O have been able to run it successfully on mine. I'll
definitely take a look.

On Thu, Jul 19, 2012 at 8:53 AM, Sharon Ben-Asher
<Sharon.Ben-Asher at avg.com>wrote:

> Hi,
>
> Today, I tried to work with v2.2.0:
> followed the instructions as close as I can, the build was successful, as
> far as I could see.
> However, when running in the device, encountered this problem:
> Cannot load library: reloc_library[1289]:   157 cannot locate
> '_ZNSt9basic_iosIcSt11char_traitsIcEE4initEPSt15basic_streambufIcS1_E'...
>
> Thanks,
> Sharon.
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120719/5dbb5b2d/attachment.htm 

From victor at iso3103.net  Thu Jul 19 12:31:33 2012
From: victor at iso3103.net (Victor Perron)
Date: Thu, 19 Jul 2012 12:31:33 +0200
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
In-Reply-To: <CAF9FejHxad-UXFthBfZcRtU-MY_c9883cymGvGvrOnH8_ccq=w@mail.gmail.com>
References: <671E27F33F2F25499F78A4AFA494E63B384D6585@ex10czdmb01>
	<CAF9FejHxad-UXFthBfZcRtU-MY_c9883cymGvGvrOnH8_ccq=w@mail.gmail.com>
Message-ID: <CAF9FejHJg_15+BGMbLotHxtXtcNCLPm5JCNeiy23VMwV8oqWCA@mail.gmail.com>

Hello,

I found the issue. I mismatched a few command line switches when writing
the wiki page; I suspect I got confused with all the configuration tweaks !
I'll update the wiki properly in the day, as soon as I tested a few more
things. It may also help for libzmq 3.x, though that's not for sure.

I uploaded a compiled version of the library (zeromq2.2 + jzmq master) on a
file hosting service, could you report if it works ?
Link : http://www9.zippyshare.com/v/83273893/file.html

Just has been tested on a SGSIII, communicates with a PC via TCP transport
(with proper android.permission.INTERNET)

On Thu, Jul 19, 2012 at 10:03 AM, Victor Perron <victor at iso3103.net> wrote:

> This is awkward. O have been able to run it successfully on mine. I'll
> definitely take a look.
>
>
> On Thu, Jul 19, 2012 at 8:53 AM, Sharon Ben-Asher <
> Sharon.Ben-Asher at avg.com> wrote:
>
>> Hi,
>>
>> Today, I tried to work with v2.2.0:
>> followed the instructions as close as I can, the build was successful, as
>> far as I could see.
>> However, when running in the device, encountered this problem:
>> Cannot load library: reloc_library[1289]:   157 cannot locate
>> '_ZNSt9basic_iosIcSt11char_traitsIcEE4initEPSt15basic_streambufIcS1_E'...
>>
>> Thanks,
>> Sharon.
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
>
> --
> Victor
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120719/64e54a8c/attachment.htm 

From steven.mccoy at miru.hk  Thu Jul 19 15:12:14 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Thu, 19 Jul 2012 09:12:14 -0400
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
	<CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
	<513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>
Message-ID: <CAGFWLigThYBU0gmU1RPRzyNRUv4SgnJA5Z4_3xSArwyJixmYwQ@mail.gmail.com>

Both are on the latest version, 5.1.118.  The fixes for OS X are in trunk
which is currently wearing the version number 5.2.119.

-- 
Steve-o

On 18 July 2012 17:31, Ron Elliott <bluerocketdelivery at gmail.com> wrote:

> Does version 3.x use the same version of openpgm that compiles on mac?
> Looks like it's now just easier to upgrade if this issue doesn't exist in
> v3.
>
> Thanks for your response
>
> On Jul 17, 2012, at 12:27 PM, Steven McCoy wrote:
>
> On 17 July 2012 02:22, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
>
>> Using the downloaded stable version 2.2 code I cannot get ZMQ to compile
>> with the --with-pgm option. The project is too far along to make the switch
>> to v3 so that's not an option atm. Any other solutions or is ZMQ dead for
>> me?
>>
>>
> It's a known issue, I have to upgrade my Mac mini from 10.6 to 10.7 to
> verify fixing the issue upstream in OpenPGM.
>
> --
> Steve-o
>  _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120719/52c41d7f/attachment.htm 

From victor at iso3103.net  Thu Jul 19 17:32:30 2012
From: victor at iso3103.net (Victor Perron)
Date: Thu, 19 Jul 2012 17:32:30 +0200
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
In-Reply-To: <CAF9FejHJg_15+BGMbLotHxtXtcNCLPm5JCNeiy23VMwV8oqWCA@mail.gmail.com>
References: <671E27F33F2F25499F78A4AFA494E63B384D6585@ex10czdmb01>
	<CAF9FejHxad-UXFthBfZcRtU-MY_c9883cymGvGvrOnH8_ccq=w@mail.gmail.com>
	<CAF9FejHJg_15+BGMbLotHxtXtcNCLPm5JCNeiy23VMwV8oqWCA@mail.gmail.com>
Message-ID: <CAF9FejFTabfGkk1j4_VeBXG1UpCo_kJFd7RwG+1cUmuNi_BHrw@mail.gmail.com>

Hello there,

I updated the wiki about this. It should work fine with ZeroMQ both 2.2 and
3.x now; I have tested them on my SGSIII, and everything seems OK.
The new method (that was actually used in my build scripts) statically
embeds the missing definitions from, in particular, libstdc++ into
libjzmq.so.
That inflates the binary size a lot (2.4 or 2.7 MBs unstripped for 2.2 and
3.x) but this can be adressed by specifying -s or running
arm-linux-androideabi-strip --strip-all on the binary. The resulting
library still works and has < 650 Kb footprint.
Please report any further issue.

Regards,

On Thu, Jul 19, 2012 at 12:31 PM, Victor Perron <victor at iso3103.net> wrote:

> Hello,
>
> I found the issue. I mismatched a few command line switches when writing
> the wiki page; I suspect I got confused with all the configuration tweaks !
> I'll update the wiki properly in the day, as soon as I tested a few more
> things. It may also help for libzmq 3.x, though that's not for sure.
>
> I uploaded a compiled version of the library (zeromq2.2 + jzmq master) on
> a file hosting service, could you report if it works ?
> Link : http://www9.zippyshare.com/v/83273893/file.html
>
> Just has been tested on a SGSIII, communicates with a PC via TCP transport
> (with proper android.permission.INTERNET)
>
>
> On Thu, Jul 19, 2012 at 10:03 AM, Victor Perron <victor at iso3103.net>wrote:
>
>> This is awkward. O have been able to run it successfully on mine. I'll
>> definitely take a look.
>>
>>
>> On Thu, Jul 19, 2012 at 8:53 AM, Sharon Ben-Asher <
>> Sharon.Ben-Asher at avg.com> wrote:
>>
>>> Hi,
>>>
>>> Today, I tried to work with v2.2.0:
>>> followed the instructions as close as I can, the build was successful,
>>> as far as I could see.
>>> However, when running in the device, encountered this problem:
>>> Cannot load library: reloc_library[1289]:   157 cannot locate
>>> '_ZNSt9basic_iosIcSt11char_traitsIcEE4initEPSt15basic_streambufIcS1_E'...
>>>
>>> Thanks,
>>> Sharon.
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>>
>>
>>
>> --
>> Victor
>>
>>
>
>
> --
> Victor
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120719/49be8ce5/attachment.htm 

From Aaron.Viviano at quadTechWorld.com  Fri Jul 20 00:31:52 2012
From: Aaron.Viviano at quadTechWorld.com (Viviano, Aaron)
Date: Thu, 19 Jul 2012 17:31:52 -0500
Subject: [zeromq-dev] Bad Pointer Exception with ZMQ 2.2.X
Message-ID: <A57FD5D0714DD84B85FB6A5A2D0BE550280A3D6127@SXEXMB3.qg.com>

Hello ZeroMQ Devs,
I'm currently running into an issue with ZMQ, using CLRZMQ (Win7, VS2010), where it is crashing in the encoder::get_data function at this line:

memcpy (buffer + pos, write_pos, to_copy);

The variable write_pos would equal 0xfeeefeee, which according to what I've researched, means that the pointer was last touched by Window's HeapFree function [1]. Looking into the code I've found that this variable is set with the address from the msg_content_t::*data variable. This is then cleaned up in zmq_msg_close:

if (content->ffn)
    content->ffn (content->data, content->hint);
free (content);

I've attempted to add a function call in zmq::encoder_t::message_ready to a function called zmq_CheckForFreedData to try and fix the problem.

bool zmq_CheckForFreedData(zmq_msg_t *msg_)
{
       if (msg_->content == (zmq::msg_content_t*) ZMQ_VSM)
        return false;
       if (msg_->content == (zmq::msg_content_t*) ZMQ_DELIMITER)
        return false;

       if(((zmq::msg_content_t*) msg_->content)->data == (void *)0xfeeefeee)
       {
              return true;
       }
       return false;
}

...
if (!source || !source->read (&in_progress))
{
       zmq_msg_init (&in_progress);
       return false;
}
else if(zmq_CheckForFreedData(&in_progress))
{
       zmq_msg_init (&in_progress);
       return false;
}
...

While this prevents the crash I'm now seeing my  program lock up. At the moment I don't have a simple test program for you, however, until I have the chance to put one together, what would be the best place to start looking for why this error is happening in the code?

Thank you,
Aaron


1.       http://stackoverflow.com/questions/127386/in-visual-studio-c-what-are-the-memory-allocation-representations
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120719/3e0cac1f/attachment.htm 

From fungust at tintedshadow.com  Fri Jul 20 01:07:21 2012
From: fungust at tintedshadow.com (Feng Yi Chan)
Date: Fri, 20 Jul 2012 07:07:21 +0800
Subject: [zeromq-dev] zeroMQ with SDP and Python on Infiniband
Message-ID: <50089329.3060606@tintedshadow.com>

Hi,

I have been trying to get libsdp.so to wrap a simple ZMQ REQ-REP socket 
in python with the following commands on two separate boxes connected 
via Infiniband.

All attempts thus far at pre-loading libsdp.so have not resulted in a 
successful connection via SDP sockets.

Instead, the ZeroMQ messages are passed via tcp, which results in poorer 
throughput on the fabric than possible with SDP.

Is there something I am doing wrong? The code and run commands are 
below. The environment and libsdp.so have been tested and shown to 
function via SDP using the iperf benchmark program. 192.168.11.11 
corresponds to the host infiniband IPoIB address.

Thank you.

LD_PRELOAD=/usr/lib64/libsdp.so LIBSDP_CONFIG_FILE=/etc/libsdp.conf 
python2.7 server.py
LD_PRELOAD=/usr/lib64/libsdp.so LIBSDP_CONFIG_FILE=/etc/libsdp.conf 
python2.7 client.py

Server code:
import zmq

context = zmq.Context()
reply = context.socket(zmq.REP)
reply.bind('tcp://192.168.11.11:5001')

while True:
     msg=reply.recv()
     reply.send(msg)

Client Code:
import zmq

context = zmq.Context()
request = context.socket(zmq.REQ)
request.connect('tcp://192.168.11.11:5001')

while True:
     request.send('hello')
     z = request.recv()

From hurtonm at gmail.com  Fri Jul 20 12:36:43 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Fri, 20 Jul 2012 12:36:43 +0200
Subject: [zeromq-dev] Bad Pointer Exception with ZMQ 2.2.X
In-Reply-To: <A57FD5D0714DD84B85FB6A5A2D0BE550280A3D6127@SXEXMB3.qg.com>
References: <A57FD5D0714DD84B85FB6A5A2D0BE550280A3D6127@SXEXMB3.qg.com>
Message-ID: <CALdZE=umpa9HCFS-38rYrQ1Q8uNbc7zx3ZdCcXMvAcVui_WrLA@mail.gmail.com>

Is this reproducible?

- Martin

On Fri, Jul 20, 2012 at 12:31 AM, Viviano, Aaron
<Aaron.Viviano at quadtechworld.com> wrote:
> Hello ZeroMQ Devs,
>
> I?m currently running into an issue with ZMQ, using CLRZMQ (Win7, VS2010),
> where it is crashing in the encoder::get_data function at this line:
>
>
>
> memcpy (buffer + pos, write_pos, to_copy);
>
>
>
> The variable write_pos would equal 0xfeeefeee, which according to what I?ve
> researched, means that the pointer was last touched by Window?s HeapFree
> function [1]. Looking into the code I?ve found that this variable is set
> with the address from the msg_content_t::*data variable. This is then
> cleaned up in zmq_msg_close:
>
>
>
> if (content->ffn)
>
>     content->ffn (content->data, content->hint);
>
> free (content);
>
>
>
> I?ve attempted to add a function call in zmq::encoder_t::message_ready to a
> function called zmq_CheckForFreedData to try and fix the problem.
>
>
>
> bool zmq_CheckForFreedData(zmq_msg_t *msg_)
>
> {
>
>        if (msg_->content == (zmq::msg_content_t*) ZMQ_VSM)
>
>         return false;
>
>        if (msg_->content == (zmq::msg_content_t*) ZMQ_DELIMITER)
>
>         return false;
>
>
>
>        if(((zmq::msg_content_t*) msg_->content)->data == (void *)0xfeeefeee)
>
>        {
>
>               return true;
>
>        }
>
>        return false;
>
> }
>
>
>
> ?
>
> if (!source || !source->read (&in_progress))
>
> {
>
>        zmq_msg_init (&in_progress);
>
>        return false;
>
> }
>
> else if(zmq_CheckForFreedData(&in_progress))
>
> {
>
>        zmq_msg_init (&in_progress);
>
>        return false;
>
> }
>
> ?
>
>
>
> While this prevents the crash I?m now seeing my  program lock up. At the
> moment I don?t have a simple test program for you, however, until I have the
> chance to put one together, what would be the best place to start looking
> for why this error is happening in the code?
>
>
>
> Thank you,
>
> Aaron
>
>
>
> 1.
> http://stackoverflow.com/questions/127386/in-visual-studio-c-what-are-the-memory-allocation-representations
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From Aaron.Viviano at quadTechWorld.com  Fri Jul 20 18:28:40 2012
From: Aaron.Viviano at quadTechWorld.com (Viviano, Aaron)
Date: Fri, 20 Jul 2012 11:28:40 -0500
Subject: [zeromq-dev] Bad Pointer Exception with ZMQ 2.2.X
In-Reply-To: <A57FD5D0714DD84B85FB6A5A2D0BE550280A3D6127@SXEXMB3.qg.com>
References: <A57FD5D0714DD84B85FB6A5A2D0BE550280A3D6127@SXEXMB3.qg.com>
Message-ID: <A57FD5D0714DD84B85FB6A5A2D0BE550280A3D6129@SXEXMB3.qg.com>

Hello ZMQ Devs,
I found the issue and it is with CLRZMQ and not LIBZMQ. Turns out that CLRZMQ was using the same object for both sending and receiving. When sending messages at a high enough rate of speed CLRZMQ was deleting the message at the same time that LIBZMQ was using it.

For those curious the issue is with "private IntPtr _msg;" in Socket.cs. Remove that variable and put unique instances in the send, forward, and receive functions that previously used the _msg variable.

My apologies for not finding that the flaw was outside LIBZMQ earlier. I will contact the CLRZMQ Devs about this.

Thank you,
Aaron Viviano
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120720/8effc151/attachment.htm 

From tw at dtex.org  Sat Jul 21 17:03:49 2012
From: tw at dtex.org (Toralf Wittner)
Date: Sat, 21 Jul 2012 17:03:49 +0200
Subject: [zeromq-dev] zmq_term can not safely be re-entered with pgm
	transport
Message-ID: <CAObngsO5EPej02bUXzz4Ra-Qao0BBO8iKAuwYe2zKM_DqtGS9A@mail.gmail.com>

According to the man page of zmq_term, it should be safe to re-enter
zmq_term if it previously returned with EINTR. This is actually not
true if used with OpenPGM. zmq_term invokes pgm_shutdown which is
reading and modifying the static variable pgm_ref_count and will
return FALSE on second invocation which makes zmq_term abort with
"Assertion failed: false (zmq.cpp:185)". This affects version 2.2 as
well as 3.2.

-- Toralf

From steven.mccoy at miru.hk  Sun Jul 22 05:50:38 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Sat, 21 Jul 2012 23:50:38 -0400
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <CAGFWLigThYBU0gmU1RPRzyNRUv4SgnJA5Z4_3xSArwyJixmYwQ@mail.gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
	<CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
	<513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>
	<CAGFWLigThYBU0gmU1RPRzyNRUv4SgnJA5Z4_3xSArwyJixmYwQ@mail.gmail.com>
Message-ID: <CAGFWLii=eKLE_mV0g689CYdSst1M8Wx+g0UimDu+gbV+-MVHaA@mail.gmail.com>

I have OS X 10.7 up and see the error whilst building with gcc.

sockaddr.c: In function 'pgm_sockaddr_pktinfo':
sockaddr.c:485: error: 'IPV6_PKTINFO' undeclared (first use in this
function)
sockaddr.c:485: error: (Each undeclared identifier is reported only once
sockaddr.c:485: error: for each function it appears in.)
make: *** [libpgm_noinst_la-sockaddr.lo] Error 1

-- 
Steve-o

On 19 July 2012 09:12, Steven McCoy <steven.mccoy at miru.hk> wrote:

> Both are on the latest version, 5.1.118.  The fixes for OS X are in trunk
> which is currently wearing the version number 5.2.119.
>
> --
> Steve-o
>
>
> On 18 July 2012 17:31, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
>
>> Does version 3.x use the same version of openpgm that compiles on mac?
>> Looks like it's now just easier to upgrade if this issue doesn't exist in
>> v3.
>>
>> Thanks for your response
>>
>> On Jul 17, 2012, at 12:27 PM, Steven McCoy wrote:
>>
>> On 17 July 2012 02:22, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
>>
>>> Using the downloaded stable version 2.2 code I cannot get ZMQ to compile
>>> with the --with-pgm option. The project is too far along to make the switch
>>> to v3 so that's not an option atm. Any other solutions or is ZMQ dead for
>>> me?
>>>
>>>
>> It's a known issue, I have to upgrade my Mac mini from 10.6 to 10.7 to
>> verify fixing the issue upstream in OpenPGM.
>>
>> --
>> Steve-o
>>  _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120721/e86de707/attachment.htm 

From steven.mccoy at miru.hk  Sun Jul 22 07:37:52 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Sun, 22 Jul 2012 01:37:52 -0400
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <CAGFWLii=eKLE_mV0g689CYdSst1M8Wx+g0UimDu+gbV+-MVHaA@mail.gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
	<CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
	<513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>
	<CAGFWLigThYBU0gmU1RPRzyNRUv4SgnJA5Z4_3xSArwyJixmYwQ@mail.gmail.com>
	<CAGFWLii=eKLE_mV0g689CYdSst1M8Wx+g0UimDu+gbV+-MVHaA@mail.gmail.com>
Message-ID: <CAGFWLijnB2NJk5G0K4JW+ERoG0uJbwDP3Ei13to4_kK_=Fg9pw@mail.gmail.com>

Try this build:

http://openpgm.googlecode.com/files/libpgm-5.2.119~dfsg.tar.gz

-- 
Steve-o

On 21 July 2012 23:50, Steven McCoy <steven.mccoy at miru.hk> wrote:

> I have OS X 10.7 up and see the error whilst building with gcc.
>
> sockaddr.c: In function 'pgm_sockaddr_pktinfo':
> sockaddr.c:485: error: 'IPV6_PKTINFO' undeclared (first use in this
> function)
> sockaddr.c:485: error: (Each undeclared identifier is reported only once
> sockaddr.c:485: error: for each function it appears in.)
> make: *** [libpgm_noinst_la-sockaddr.lo] Error 1
>
> --
> Steve-o
>
>
> On 19 July 2012 09:12, Steven McCoy <steven.mccoy at miru.hk> wrote:
>
>> Both are on the latest version, 5.1.118.  The fixes for OS X are in trunk
>> which is currently wearing the version number 5.2.119.
>>
>> --
>> Steve-o
>>
>>
>> On 18 July 2012 17:31, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
>>
>>> Does version 3.x use the same version of openpgm that compiles on mac?
>>> Looks like it's now just easier to upgrade if this issue doesn't exist in
>>> v3.
>>>
>>> Thanks for your response
>>>
>>> On Jul 17, 2012, at 12:27 PM, Steven McCoy wrote:
>>>
>>> On 17 July 2012 02:22, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
>>>
>>>> Using the downloaded stable version 2.2 code I cannot get ZMQ to
>>>> compile with the --with-pgm option. The project is too far along to make
>>>> the switch to v3 so that's not an option atm. Any other solutions or is ZMQ
>>>> dead for me?
>>>>
>>>>
>>> It's a known issue, I have to upgrade my Mac mini from 10.6 to 10.7 to
>>> verify fixing the issue upstream in OpenPGM.
>>>
>>> --
>>> Steve-o
>>>  _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>>>
>>>
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120722/52299f37/attachment.htm 

From varokan at movingsatellites.com  Fri Jul 20 19:36:57 2012
From: varokan at movingsatellites.com (varokan at movingsatellites.com)
Date: Fri, 20 Jul 2012 10:36:57 -0700
Subject: [zeromq-dev] ZMQ and TLS ( again )
Message-ID: <efa3e420fad9f2c3f4838305549a109a.squirrel@webmail.movingsatellites.com>

Just a quick question regarding encryption.

I read that there are considerations to leverage tcpcrypt or similar for
opportunistci encryption but I also found Wes Youngs GnuTLS implementation
in ZeroMQ itself.
https://github.com/wesyoung/libzmq

I rather prefer the integrated approach Wes Yuong used through
"zmq_setsockopt ( socket, ZMQ_TLS, "true", 5 )" as it is easier to use,
better integrated into the ZeroMQ API, and overall much cleaner from a
users point of view.

What are the chances to integrate this into the latest ZeroMQ code base ?
It seems to me that the impact would be minimal and the benefits would be
outstanding.

If the documentation points out exactly when and how to use TLS, then I
think the concerns about unsupported communication types can be mitigated.
Also if this can be enabled during compile time through a switch to the
configure script it would not interfere with the other core features.

My last point is that the changes which I saw on git seem minimal, which
should be good news to any one concerned about code-bloat.

Encryption is or should be an inherent capability of sockets and socket
libraries in this day and age.

Please consider adding TLS encryption to ZeroMQ, if only as alpha for now.

Thanks,

Varol



From bluerocketdelivery at gmail.com  Sun Jul 22 06:43:35 2012
From: bluerocketdelivery at gmail.com (Ron Elliott)
Date: Sat, 21 Jul 2012 21:43:35 -0700
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <CAGFWLii=eKLE_mV0g689CYdSst1M8Wx+g0UimDu+gbV+-MVHaA@mail.gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
	<CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
	<513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>
	<CAGFWLigThYBU0gmU1RPRzyNRUv4SgnJA5Z4_3xSArwyJixmYwQ@mail.gmail.com>
	<CAGFWLii=eKLE_mV0g689CYdSst1M8Wx+g0UimDu+gbV+-MVHaA@mail.gmail.com>
Message-ID: <361CE0FF-66F8-4BA6-B406-0434B9EADF65@gmail.com>

That should be remedied by adding -D__APPLE_USE_RFC_2292 to the compile flags, see this bug report:

http://code.google.com/p/openpgm/issues/detail?id=22

I have 3-x compiling now, but progress getting my code back to a testable state is slow so I'm still not 100% sure that it is working correctly but at least the compiler isn't complaining.

Worth noting, if you are trying to build openpgm and keep getting a "waiting for lock release on file ?." error and you happen to have the source on an exFAT partition, move it off of there to get it to compile successfully.

On Jul 21, 2012, at 8:50 PM, Steven McCoy wrote:

> I have OS X 10.7 up and see the error whilst building with gcc.  
> 
> sockaddr.c: In function 'pgm_sockaddr_pktinfo':
> sockaddr.c:485: error: 'IPV6_PKTINFO' undeclared (first use in this function)
> sockaddr.c:485: error: (Each undeclared identifier is reported only once
> sockaddr.c:485: error: for each function it appears in.)
> make: *** [libpgm_noinst_la-sockaddr.lo] Error 1
> 
> -- 
> Steve-o
> 
> On 19 July 2012 09:12, Steven McCoy <steven.mccoy at miru.hk> wrote:
> Both are on the latest version, 5.1.118.  The fixes for OS X are in trunk which is currently wearing the version number 5.2.119.
> 
> -- 
> Steve-o
> 
> 
> On 18 July 2012 17:31, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
> Does version 3.x use the same version of openpgm that compiles on mac? Looks like it's now just easier to upgrade if this issue doesn't exist in v3.
> 
> Thanks for your response
> 
> On Jul 17, 2012, at 12:27 PM, Steven McCoy wrote:
> 
>> On 17 July 2012 02:22, Ron Elliott <bluerocketdelivery at gmail.com> wrote:
>> Using the downloaded stable version 2.2 code I cannot get ZMQ to compile with the --with-pgm option. The project is too far along to make the switch to v3 so that's not an option atm. Any other solutions or is ZMQ dead for me?
>> 
>> 
>> It's a known issue, I have to upgrade my Mac mini from 10.6 to 10.7 to verify fixing the issue upstream in OpenPGM.
>> 
>> -- 
>> Steve-o 
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 
> 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 
> 
> 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120721/02871458/attachment.htm 

From ph at imatix.com  Sun Jul 22 13:39:10 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Sun, 22 Jul 2012 13:39:10 +0200
Subject: [zeromq-dev] zmq_term can not safely be re-entered with pgm
	transport
In-Reply-To: <CAObngsO5EPej02bUXzz4Ra-Qao0BBO8iKAuwYe2zKM_DqtGS9A@mail.gmail.com>
References: <CAObngsO5EPej02bUXzz4Ra-Qao0BBO8iKAuwYe2zKM_DqtGS9A@mail.gmail.com>
Message-ID: <CADL5_sjm40ES1dtHeHykdVUiTXKRQ-xDMOWrB9oUKzc-yjeShg@mail.gmail.com>

Hi Toralf,

Thanks for spotting this. Would you like to create an issue for this
in our issue tracker?

And, if you have time, fix this and send us pull requests?

Cheers
Pieter

On Sat, Jul 21, 2012 at 5:03 PM, Toralf Wittner <tw at dtex.org> wrote:
> According to the man page of zmq_term, it should be safe to re-enter
> zmq_term if it previously returned with EINTR. This is actually not
> true if used with OpenPGM. zmq_term invokes pgm_shutdown which is
> reading and modifying the static variable pgm_ref_count and will
> return FALSE on second invocation which makes zmq_term abort with
> "Assertion failed: false (zmq.cpp:185)". This affects version 2.2 as
> well as 3.2.
>
> -- Toralf
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From steven.mccoy at miru.hk  Sun Jul 22 16:35:48 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Sun, 22 Jul 2012 10:35:48 -0400
Subject: [zeromq-dev] OS 10.7 Cannot Build with --with-pgm
In-Reply-To: <361CE0FF-66F8-4BA6-B406-0434B9EADF65@gmail.com>
References: <ECFC30B5-F2E5-417B-BBAA-A3FB124AA3CB@gmail.com>
	<CAGFWLij5dpgQx8_2ERuLkiDRP=mtuejSAgMx2_KbO=BK5fMHRw@mail.gmail.com>
	<513250A7-472F-4FB1-AC41-13883E1AEA13@gmail.com>
	<CAGFWLigThYBU0gmU1RPRzyNRUv4SgnJA5Z4_3xSArwyJixmYwQ@mail.gmail.com>
	<CAGFWLii=eKLE_mV0g689CYdSst1M8Wx+g0UimDu+gbV+-MVHaA@mail.gmail.com>
	<361CE0FF-66F8-4BA6-B406-0434B9EADF65@gmail.com>
Message-ID: <CAGFWLijZ7Nh_Z5jnqf4Kk2ZUgwmsr8KqsXp7+nN7cvhh4q1dWA@mail.gmail.com>

On 22 July 2012 00:43, Ron Elliott <bluerocketdelivery at gmail.com> wrote:

> That should be remedied by adding -D__APPLE_USE_RFC_2292 to the compile
> flags, see this bug report:
>
>
I actually read the headers this time:

*/usr/include/netinet6/in6.h*:
/* RFC 3542 define the following socket options in a manner incompatible
 * with RFC 2292:
 *   IPV6_PKTINFO
 *   IPV6_HOPLIMIT
 *   IPV6_NEXTHOP
 *   IPV6_HOPOPTS
 *   IPV6_DSTOPTS
 *   IPV6_RTHDR
 *
 * To use the new IPv6 Sockets options introduced by RFC 3542
 * the constant __APPLE_USE_RFC_3542 must be defined before
 * including <netinet/in.h>
 *
 * To use the old IPv6 Sockets options from RFC 2292
 * the constant __APPLE_USE_RFC_2292 must be defined before
 * including <netinet/in.h>
 *
 * Note that eventually RFC 3542 is going to be the
 * default and RFC 2292 will be obsolete.
 */

The RFC 2292 version of IPV6_PKTINFO:

#define IPV6_2292PKTINFO        19 /* bool; send/recv if, src/dst addr */

The RFC 3542 version:

#define IPV6_RECVPKTINFO        61 /* bool; recv if, dst addr
                                    * APPLE: Value purposely different than
FreeBSD (36) to avoid
                                    * collision with definition of
IPV6_TCLASS in previous
                                    * darwin implementations */

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120722/97ffd4d1/attachment.htm 

From steven.mccoy at miru.hk  Sun Jul 22 16:46:31 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Sun, 22 Jul 2012 10:46:31 -0400
Subject: [zeromq-dev] zmq_term can not safely be re-entered with pgm
	transport
In-Reply-To: <CADL5_sjm40ES1dtHeHykdVUiTXKRQ-xDMOWrB9oUKzc-yjeShg@mail.gmail.com>
References: <CAObngsO5EPej02bUXzz4Ra-Qao0BBO8iKAuwYe2zKM_DqtGS9A@mail.gmail.com>
	<CADL5_sjm40ES1dtHeHykdVUiTXKRQ-xDMOWrB9oUKzc-yjeShg@mail.gmail.com>
Message-ID: <CAGFWLihSzj8h9SpWO=z8x=Wey4XmpLhyiq-VUC6ia4UmkxWQtw@mail.gmail.com>

Plus its a similar problem with Windows.  Pre-emptive patch attached.

-- 
Steve-o

On 22 July 2012 07:39, Pieter Hintjens <ph at imatix.com> wrote:

> Hi Toralf,
>
> Thanks for spotting this. Would you like to create an issue for this
> in our issue tracker?
>
> And, if you have time, fix this and send us pull requests?
>
> Cheers
> Pieter
>
> On Sat, Jul 21, 2012 at 5:03 PM, Toralf Wittner <tw at dtex.org> wrote:
> > According to the man page of zmq_term, it should be safe to re-enter
> > zmq_term if it previously returned with EINTR. This is actually not
> > true if used with OpenPGM. zmq_term invokes pgm_shutdown which is
> > reading and modifying the static variable pgm_ref_count and will
> > return FALSE on second invocation which makes zmq_term abort with
> > "Assertion failed: false (zmq.cpp:185)". This affects version 2.2 as
> > well as 3.2.
> >
> > -- Toralf
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120722/5747e254/attachment.htm 
-------------- next part --------------
--- zmq.cpp.orig        2012-07-22 10:42:24.000000000 -0400
+++ zmq.cpp 2012-07-22 10:44:36.000000000 -0400
@@ -292,17 +292,20 @@
     int rc = ((zmq::ctx_t*) ctx_)->terminate ();
     int en = errno;

+    //  Shut down only if termination was not interrupted by a signal.
+    if (!rc && en == EINTR) {
 #ifdef ZMQ_HAVE_WINDOWS
-    //  On Windows, uninitialise socket layer.
-    rc = WSACleanup ();
-    wsa_assert (rc != SOCKET_ERROR);
+        //  On Windows, uninitialise socket layer.
+        rc = WSACleanup ();
+        wsa_assert (rc != SOCKET_ERROR);
 #endif

 #if defined ZMQ_HAVE_OPENPGM
-    //  Shut down the OpenPGM library.
-    if (pgm_shutdown () != TRUE)
-        zmq_assert (false);
+        //  Shut down the OpenPGM library.
+        if (pgm_shutdown () != TRUE)
+            zmq_assert (false);
 #endif
+    }

     errno = en;
     return rc;

From steven.mccoy at miru.hk  Sun Jul 22 17:24:18 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Sun, 22 Jul 2012 11:24:18 -0400
Subject: [zeromq-dev] zmq_term can not safely be re-entered with pgm
	transport
In-Reply-To: <CAGFWLihSzj8h9SpWO=z8x=Wey4XmpLhyiq-VUC6ia4UmkxWQtw@mail.gmail.com>
References: <CAObngsO5EPej02bUXzz4Ra-Qao0BBO8iKAuwYe2zKM_DqtGS9A@mail.gmail.com>
	<CADL5_sjm40ES1dtHeHykdVUiTXKRQ-xDMOWrB9oUKzc-yjeShg@mail.gmail.com>
	<CAGFWLihSzj8h9SpWO=z8x=Wey4XmpLhyiq-VUC6ia4UmkxWQtw@mail.gmail.com>
Message-ID: <CAGFWLijv3MAcGGfZeNfZ1UC3gu2wR5xTf-8xchqq5Wx3SR1r=w@mail.gmail.com>

Almost right, Toralf waiting on the Jira case to fix.

https://zeromq.jira.com

-if (!rc && en == EINTR) {

+if (!rc || en != EINTR) {

-- 
Steve-o

On 22 July 2012 10:46, Steven McCoy <steven.mccoy at miru.hk> wrote:

> Plus its a similar problem with Windows.  Pre-emptive patch attached.
>
> --
> Steve-o
>
>
> On 22 July 2012 07:39, Pieter Hintjens <ph at imatix.com> wrote:
>
>> Hi Toralf,
>>
>> Thanks for spotting this. Would you like to create an issue for this
>> in our issue tracker?
>>
>> And, if you have time, fix this and send us pull requests?
>>
>> Cheers
>> Pieter
>>
>> On Sat, Jul 21, 2012 at 5:03 PM, Toralf Wittner <tw at dtex.org> wrote:
>> > According to the man page of zmq_term, it should be safe to re-enter
>> > zmq_term if it previously returned with EINTR. This is actually not
>> > true if used with OpenPGM. zmq_term invokes pgm_shutdown which is
>> > reading and modifying the static variable pgm_ref_count and will
>> > return FALSE on second invocation which makes zmq_term abort with
>> > "Assertion failed: false (zmq.cpp:185)". This affects version 2.2 as
>> > well as 3.2.
>> >
>> > -- Toralf
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120722/e9069491/attachment.htm 

From wes at barely3am.com  Mon Jul 23 03:34:46 2012
From: wes at barely3am.com (Wes Young)
Date: Sun, 22 Jul 2012 21:34:46 -0400
Subject: [zeromq-dev] ZMQ and TLS ( again )
In-Reply-To: <efa3e420fad9f2c3f4838305549a109a.squirrel@webmail.movingsatellites.com>
References: <efa3e420fad9f2c3f4838305549a109a.squirrel@webmail.movingsatellites.com>
Message-ID: <DDB14870-A8C1-4D26-8A01-D80E640B2B19@barely3am.com>

> I rather prefer the integrated approach Wes Yuong used through
> "zmq_setsockopt ( socket, ZMQ_TLS, "true", 5 )" as it is easier to use,
> better integrated into the ZeroMQ API, and overall much cleaner from a
> users point of view.

:)

> What are the chances to integrate this into the latest ZeroMQ code base ?
> It seems to me that the impact would be minimal and the benefits would be
> outstanding.

fwiw: [at the time] i was waiting for 3.x to settle down a bit before diving back into this.

> If the documentation points out exactly when and how to use TLS, then I
> think the concerns about unsupported communication types can be mitigated.
> Also if this can be enabled during compile time through a switch to the
> configure script it would not interfere with the other core features.
> 
> My last point is that the changes which I saw on git seem minimal, which
> should be good news to any one concerned about code-bloat.

we were trying to do it as a compile time argument for starters, and what we started figuring out (to be tested) after reading through [just about all of] libgnutls is that we might even be able to get away with "anonymous" TLS at the zmq level and leave the "Identity / auth" up to the higher levels of the protocol itself (being implemented in our testing as protobuf atm).

> Encryption is or should be an inherent capability of sockets and socket
> libraries in this day and age.

while i don't disagree, and i think the solution will be a rather elegant one (i hope :x); it's a tough problem to solve correctly. some of which is determined on how the low-level stack was/is implemented (the way libzmq works and the way gnutls works)?

> Please consider adding TLS encryption to ZeroMQ, if only as alpha for now.

the original branch i wrote should be considered alpha in and of itself. i hadn't gotten much feedback other than "that's a neat idea" yet. this kind of crypto (at this level) could prove to make zmq unstable, which is why even in the branch you have to add it at compile time as a ./configure flag.

crypto is hard; gnutls is worse; and if you do it wrong you're going to screw a *lot* of people in the process. which is why i'm guessing i haven't gotten a lot of other eyeballs on it just yet either :)

either way, we should have more traction on our end for this come early fall or so. we'll be getting to a point in our project where we'll start testing these kinda links at scale? one way or the other it's on our list of things where it's a requirement to be solved.

cheers,

--
Wes
wesyoung.me
collectiveintel.net

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 203 bytes
Desc: Message signed with OpenPGP using GPGMail
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120722/c2664297/attachment.pgp 

From vitaliy at toroki.com  Mon Jul 23 15:15:40 2012
From: vitaliy at toroki.com (Vitaliy Ivanov)
Date: Mon, 23 Jul 2012 16:15:40 +0300
Subject: [zeromq-dev] [PATCH] tests: getting rid of warnings when build with
	NDEBUG.
Message-ID: <1343049340.16350.16.camel@vivanov>

Hi,

ZMQ tree 2-x.
I didn't use ZMQ_ASSERT as tests are not using this macro.

Vitaliy

----------------------------------------------------------------------

>From 12e7514e9c076f35cd525fee5f2c68a1228e53f2 Mon Sep 17 00:00:00 2001
From: Vitaliy Ivanov <vitaliy at toroki.com>
Date: Mon, 23 Jul 2012 16:06:13 +0300
Subject: [PATCH] tests: getting rid of warnings when build with NDEBUG.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Correcting the following warnings when build in non asserts mode:
a.
testutil.hpp: In function ?void zmqtestutil::basic_tests(const char*, int, int)?:
testutil.hpp:114: error: unused variable ?rc?
b.
test_hwm.cpp: In function ?int main(int, char**)?:
test_hwm.cpp:47: error: unused variable ?sent?
test_hwm.cpp:60: error: unused variable ?received?
test_hwm.cpp:64: error: unused variable ?sent?

Signed-off-by: Vitaliy Ivanov <vitalivanov at gmail.com>
---
 tests/test_hwm.cpp   |   17 +++++++++--------
 tests/test_timeo.cpp |    4 ++--
 tests/testutil.hpp   |   17 +++++++++--------
 3 files changed, 20 insertions(+), 18 deletions(-)

diff --git a/tests/test_hwm.cpp b/tests/test_hwm.cpp
index d73c25f..e1fcec2 100644
--- a/tests/test_hwm.cpp
+++ b/tests/test_hwm.cpp
@@ -28,6 +28,7 @@ int main (int argc, char *argv [])
 {
     uint64_t hwm = 5;
     int linger = 0;
+    bool rc;
 
     zmq::context_t context (1);
     zmq::socket_t s1 (context, ZMQ_PULL);
@@ -44,25 +45,25 @@ int main (int argc, char *argv [])
         zmq::message_t msg (sizeof ("test") - 1);
         memcpy (msg.data (), "test", sizeof ("test") - 1);
 
-        bool sent = s2.send (msg, ZMQ_NOBLOCK);
+        rc = s2.send (msg, ZMQ_NOBLOCK);
 
         // Anything below HWM should be sent
         if (i < 5) {
-            assert (sent);
+            assert (rc);
         } else {
-            assert (!sent && errno == EAGAIN);
+            assert (!rc && errno == EAGAIN);
         }
     }
 
     // There should be now 5 messages pending, consume one
     zmq::message_t msg;
 
-    bool received = s1.recv (&msg, 0);
-    assert (received);
+    rc = s1.recv (&msg, 0);
+    assert (rc);
 
     // Now it should be possible to send one more
-    bool sent = s2.send (msg, 0);
-    assert (sent);
+    rc = s2.send (msg, 0);
+    assert (rc);
 
-	return 0;
+    return 0;
 }
diff --git a/tests/test_timeo.cpp b/tests/test_timeo.cpp
index b9cd512..251972e 100644
--- a/tests/test_timeo.cpp
+++ b/tests/test_timeo.cpp
@@ -107,7 +107,7 @@ int main (int argc, char *argv [])
     memcpy (zmq_msg_data (&msg), "12345678ABCDEFGH12345678abcdefgh", 32);
     rc = zmq_send (sc, &msg, 0);
     assert (rc == 0);
-    
+
     rc = zmq_recv (sb, &msg, 0);
     assert (rc == 0);
     assert (32 == zmq_msg_size(&msg));
@@ -120,6 +120,6 @@ int main (int argc, char *argv [])
     rc = zmq_term (ctx);
     assert (rc == 0);
 
-    return 0 ;
+    return 0;
 }
 
diff --git a/tests/testutil.hpp b/tests/testutil.hpp
index 7f182c1..550ecf0 100644
--- a/tests/testutil.hpp
+++ b/tests/testutil.hpp
@@ -30,8 +30,7 @@
 
 namespace zmqtestutil
 {
-
-    using namespace std ;
+    using namespace std;
 
     typedef std::pair <zmq::socket_t*, zmq::socket_t*> socket_pair;
 
@@ -68,7 +67,7 @@ namespace zmqtestutil
         s2.send (pong, 0);
 
         //  Return received data as std::string.
-        return ret ;
+        return ret;
     }
 
     /*  Run basic tests for the given transport.
@@ -79,6 +78,8 @@ namespace zmqtestutil
     */
     void basic_tests (const char *transport_, int t1_, int t2_)
     {
+        int rc = 0;
+        string returned;
         zmq::context_t context (1);
 
         zmq::pollitem_t items [2];
@@ -88,7 +89,7 @@ namespace zmqtestutil
         const string expect ("XXX");
 
         {
-            const string returned = zmqtestutil::ping_pong (p, expect);
+            returned = zmqtestutil::ping_pong (p, expect);
             assert (expect == returned);
 
             //  Adjust socket state so that poll shows only 1 pending message.
@@ -97,7 +98,7 @@ namespace zmqtestutil
         }
 
         {
-            //  Now poll is used to singal that a message is ready to read.
+            //  Now poll is used to signal that a message is ready to read.
             zmq::message_t m1 (expect.size ());
             memcpy (m1.data (), expect.c_str (), expect.size ());
             items [0].socket = *p.first;
@@ -111,14 +112,14 @@ namespace zmqtestutil
 
             p.first->send (m1, 0);
 
-            int rc = zmq::poll (&items [0], 2, -1);
+            rc = zmq::poll (&items [0], 2, -1);
             assert (rc == 1);
             assert ((items [1].revents & ZMQ_POLLIN) != 0);
 
             zmq::message_t m2;
             p.second->recv (&m2, 0);
-            const string ret ((char*) m2.data (), m2.size ());
-            assert (expect == ret);
+            returned = string((char*) m2.data (), m2.size ());
+            assert (expect == returned);
         }
 
         //  Delete sockets.
-- 
1.7.0.4




From Sharon.Ben-Asher at avg.com  Mon Jul 23 15:25:31 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Mon, 23 Jul 2012 13:25:31 +0000
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
 Android APK
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D7893@ex10czdmb01>

Victor,

I can confirm that the precompiled shared object works fine.  I will try to re-compile the source with the updated instructions.

Sharon.

-----Original Message-----
From: Victor Perron <victor at iso3103.net>
Subject: Re: [zeromq-dev] ZeroMQ for Android: Problem in building the	Android APK
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org

Hello,

I found the issue. I mismatched a few command line switches when writing the wiki page; I suspect I got confused with all the configuration tweaks !
I'll update the wiki properly in the day, as soon as I tested a few more things. It may also help for libzmq 3.x, though that's not for sure.

I uploaded a compiled version of the library (zeromq2.2 + jzmq master) on a file hosting service, could you report if it works ?
Link : http://www9.zippyshare.com/v/83273893/file.html

Just has been tested on a SGSIII, communicates with a PC via TCP transport (with proper android.permission.INTERNET)


From n1555 at yahoo.fr  Mon Jul 23 20:53:57 2012
From: n1555 at yahoo.fr (Eric)
Date: Mon, 23 Jul 2012 19:53:57 +0100 (BST)
Subject: [zeromq-dev] ZeroMQ for Android: Problem in building the
	Android APK
In-Reply-To: <671E27F33F2F25499F78A4AFA494E63B384D7893@ex10czdmb01>
Message-ID: <1343069637.8613.YahooMailClassic@web29503.mail.ird.yahoo.com>

Hi all,
I confirm that the precompiled library works too for my Acer Liquid with Android 2.2. 
Thanks!
I have compiled the library but have not tested it.



--- En date de?: Lun 23.7.12, Sharon Ben-Asher <Sharon.Ben-Asher at avg.com> a ?crit?:

De: Sharon Ben-Asher <Sharon.Ben-Asher at avg.com>
Objet: Re: [zeromq-dev] ZeroMQ for Android: Problem in building the Android APK
?: "zeromq-dev at lists.zeromq.org" <zeromq-dev at lists.zeromq.org>
Date: Lundi 23 juillet 2012, 15h25

Victor,

I can confirm that the precompiled shared object works fine.? I will try to re-compile the source with the updated instructions.

Sharon.

-----Original Message-----
From: Victor Perron <victor at iso3103.net>
Subject: Re: [zeromq-dev] ZeroMQ for Android: Problem in building the??? Android APK
To: ZeroMQ development list <zeromq-dev at lists.zeromq.org

Hello,

I found the issue. I mismatched a few command line switches when writing the wiki page; I suspect I got confused with all the configuration tweaks !
I'll update the wiki properly in the day, as soon as I tested a few more things. It may also help for libzmq 3.x, though that's not for sure.

I uploaded a compiled version of the library (zeromq2.2 + jzmq master) on a file hosting service, could you report if it works ?
Link : http://www9.zippyshare.com/v/83273893/file.html

Just has been tested on a SGSIII, communicates with a PC via TCP transport (with proper android.permission.INTERNET)

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120723/dfdc575b/attachment.htm 

From finalguy at gmail.com  Tue Jul 24 06:47:50 2012
From: finalguy at gmail.com (Ming Ji)
Date: Mon, 23 Jul 2012 23:47:50 -0500
Subject: [zeromq-dev] Question about constness
Message-ID: <500E28F6.1040608@gmail.com>

Hi all,

I am new to ZeroMQ. I am actually not quite comfortable about the 
constness in the API.

For example,

size_t zmq_msg_size(zmq_msg_t* msg);

In my opinion, it is better to do it this way:

size_t zmq_msg_size(const zmq_msg_t* msg);

because this function by definition only returns the size of the 
message. It should not change the message itself. By adding the const 
there, developers can pass addresses of const messages into this function.

Any comments?



From paolo.denti at gmail.com  Tue Jul 24 12:22:11 2012
From: paolo.denti at gmail.com (Paolo Denti)
Date: Tue, 24 Jul 2012 12:22:11 +0200
Subject: [zeromq-dev] compiling zeromq on ios
Message-ID: <35885C68-9C2A-4276-B8F7-99BE61AD634C@gmail.com>

Hi all,
i am trying to compile zeromq for ios5 and i had to change a little bit the instructions found on the wiki at page http://www.zeromq.org/build:iphone

SDK_ROOT="/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer"
export CXX="${SDK_ROOT}/usr/bin/g++"
export CC="${SDK_ROOT}/usr/bin/gcc"
./configure --disable-dependency-tracking --enable-static --disable-shared --host=arm-apple-darwin10 --with-poller=kqueue --prefix=${BUILD_DIR}

changing the xcode location and adding --with-poller=kqueue because a polling system was not found (and kqueue is included in ios)

However while configuring i get the following warnings (autoconf version is 2.61)

checking errno.h usability... no
checking errno.h presence... yes
configure: WARNING: errno.h: present but cannot be compiled
configure: WARNING: errno.h:     check for missing prerequisite headers?
configure: WARNING: errno.h: see the Autoconf documentation
configure: WARNING: errno.h:     section "Present But Cannot Be Compiled"
configure: WARNING: errno.h: proceeding with the compiler's result
configure: WARNING:     ## ------------------------------------------ ##
configure: WARNING:     ## Report this to zeromq-dev at lists.zeromq.org ##
configure: WARNING:     ## ------------------------------------------ ##
checking for errno.h... no
checking arpa/inet.h usability... no
checking arpa/inet.h presence... yes
configure: WARNING: arpa/inet.h: present but cannot be compiled
configure: WARNING: arpa/inet.h:     check for missing prerequisite headers?
configure: WARNING: arpa/inet.h: see the Autoconf documentation
configure: WARNING: arpa/inet.h:     section "Present But Cannot Be Compiled"
configure: WARNING: arpa/inet.h: proceeding with the compiler's result
configure: WARNING:     ## ------------------------------------------ ##
configure: WARNING:     ## Report this to zeromq-dev at lists.zeromq.org ##
configure: WARNING:     ## ------------------------------------------ ##
checking for arpa/inet.h... no
....

and several others
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120724/7d3148af/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3642 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120724/7d3148af/attachment.bin 

From sebastian.lauwers at gmail.com  Tue Jul 24 15:23:34 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Tue, 24 Jul 2012 15:23:34 +0200
Subject: [zeromq-dev] Question about constness
In-Reply-To: <500E28F6.1040608@gmail.com>
References: <500E28F6.1040608@gmail.com>
Message-ID: <CAESgSZqOZGD+oNrzSuNF_JjH8sbqW0aZ1--VVu67=r=Y=scG1g@mail.gmail.com>

On 24 July 2012 06:47, Ming Ji <finalguy at gmail.com> wrote:
> Hi all,

Hi there,

> I am new to ZeroMQ. I am actually not quite comfortable about the
> constness in the API.

You are not the only one. This has been discussed a few times already:

http://lists.zeromq.org/pipermail/zeromq-dev/2010-October/007392.html
http://lists.zeromq.org/pipermail/zeromq-dev/2011-May/011668.html
http://lists.zeromq.org/pipermail/zeromq-dev/2010-July/004435.html

> For example,
> size_t zmq_msg_size(const zmq_msg_t* msg);

Indeed, but it also requires all the underlying calls to be
const-compatible. This, in and of itself, is not really an issue, but
as noted in the threads cited above, is prone to breakage unless  done
in major revisions.

Checking the zeromq 3.2 RC [1], I see this still hasn't been
applied---I don't know what the project's policy would be regarding
"breaking" the 3.2 API after the first RC has been released.
Constness is used a handful of times in zmq.h.

> Any comments?

Personally, I've wrapped the zmq.hpp API with my own interface (the
project that contains this is LGPL and will be released around
September), and which adds a number of objects to the standard CPP API
(ZMQSendingSocket, ZMQReceivingSocket, ZMQPublisherSocket, etc), and
full const-correctness. I didn't feel the need to change the zeromq
API per se, but it certainly might make things easier/safer for
others.

If there is a demand for it, I could probably provide a patch for the
CPP API, or even share my interface, if asked.

Lastly, on a purely stylistic note, I prefer adding the const modifier
after the symbol it applies to. I would thus re-write your provided
example as:

size_t zmq_msg_size(zmq_msg_t const * msg);

[1]: https://github.com/zeromq/zeromq3-x/blob/master/include/zmq.h

PS: First post on mailing lists. Hip hip.
PPS: Moderators, I tried sending from my @googlemail address, which
obviously wasn't accepted by mailman. Apologies.

From aleemb at gmail.com  Tue Jul 24 12:25:17 2012
From: aleemb at gmail.com (Aleem B)
Date: Tue, 24 Jul 2012 15:25:17 +0500
Subject: [zeromq-dev] 64 bit PHP extension
Message-ID: <CA+=Uhp-c1-p9PdmqV3iZC4vVoza6FMeL7mzTgFTH_etTFCP4YA@mail.gmail.com>

The snapshot page does not offer a 64 bit PHP extension. Has anyone managed
to compile and generate an x64 PHP DLL? Would really appreciate a copy of
the file.

-- Aleem
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120724/419fb6fb/attachment.htm 

From jonas.adler at epiq.se  Tue Jul 24 16:03:33 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Tue, 24 Jul 2012 16:03:33 +0200 (CEST)
Subject: [zeromq-dev] Massive problems running ZMQ on android
Message-ID: <a469afc041f1b0e985d760584d6d0866.squirrel@mail.glesys.se>

Hello,

I'm new to ZMQ, and trying to get it to run on android. I am however,
facing issue, after issue, after issue.

I try to follow the official guide, (http://www.zeromq.org/build:android),
and have gotten to the following call on my ubuntu machine:

./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
--with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version

I get this error:

"configure: error: cannot link with -lzmq"

I have tried google for a few days now, but i cannot find any solution to
this that works for me.

Some comments on other solutions i have found, that did not work for me:

I have installed pretty much every compiler known to man, but if you have
any specific suggestions, please tell me.
I have made no alterations to the source at all, simply following the
instructions to the letter.

Has someone else had this error? Is there something that is "understood"
in the guide, that i could be missing?

Any help would be greatly appreciated!

//Jonas Adler


From victor at iso3103.net  Tue Jul 24 19:25:28 2012
From: victor at iso3103.net (Victor Perron)
Date: Tue, 24 Jul 2012 19:25:28 +0200
Subject: [zeromq-dev] Massive problems running ZMQ on android
In-Reply-To: <a469afc041f1b0e985d760584d6d0866.squirrel@mail.glesys.se>
References: <a469afc041f1b0e985d760584d6d0866.squirrel@mail.glesys.se>
Message-ID: <CAF9FejE4Usydypr0iLPyz5PFyHjjHHj6PDJBePh2vM1qLebVdg@mail.gmail.com>

Seems that you were compiling jzmq without having previously ran a 'make
install' of zeromq itself, is it ?

Check if that is the issue, best regards
Le 24 juil. 2012 16:40, <jonas.adler at epiq.se> a ?crit :

> Hello,
>
> I'm new to ZMQ, and trying to get it to run on android. I am however,
> facing issue, after issue, after issue.
>
> I try to follow the official guide, (http://www.zeromq.org/build:android),
> and have gotten to the following call on my ubuntu machine:
>
> ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
>
> I get this error:
>
> "configure: error: cannot link with -lzmq"
>
> I have tried google for a few days now, but i cannot find any solution to
> this that works for me.
>
> Some comments on other solutions i have found, that did not work for me:
>
> I have installed pretty much every compiler known to man, but if you have
> any specific suggestions, please tell me.
> I have made no alterations to the source at all, simply following the
> instructions to the letter.
>
> Has someone else had this error? Is there something that is "understood"
> in the guide, that i could be missing?
>
> Any help would be greatly appreciated!
>
> //Jonas Adler
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120724/7e60f001/attachment.htm 

From n1555 at yahoo.fr  Tue Jul 24 19:26:04 2012
From: n1555 at yahoo.fr (Eric)
Date: Tue, 24 Jul 2012 18:26:04 +0100 (BST)
Subject: [zeromq-dev] Re :  Massive problems running ZMQ on android
In-Reply-To: <a469afc041f1b0e985d760584d6d0866.squirrel@mail.glesys.se>
Message-ID: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>

Did you try to configure zmq or jzmq?The configure command you described is command line for jzmq. You need first to have set the env variable?OUTPUT_DIR=/tmp/zeromq-android?and then have generated zmq.
Then you should be able to launch configure for jzmq.
Regards.

--- En date de?: Mar 24.7.12, jonas.adler at epiq.se <jonas.adler at epiq.se> a ?crit?:

De: jonas.adler at epiq.se <jonas.adler at epiq.se>
Objet: [zeromq-dev] Massive problems running ZMQ on android
?: zeromq-dev at lists.zeromq.org
Date: Mardi 24 juillet 2012, 16h03

Hello,

I'm new to ZMQ, and trying to get it to run on android. I am however,
facing issue, after issue, after issue.

I try to follow the official guide, (http://www.zeromq.org/build:android),
and have gotten to the following call on my ubuntu machine:

./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
--with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version

I get this error:

"configure: error: cannot link with -lzmq"

I have tried google for a few days now, but i cannot find any solution to
this that works for me.

Some comments on other solutions i have found, that did not work for me:

I have installed pretty much every compiler known to man, but if you have
any specific suggestions, please tell me.
I have made no alterations to the source at all, simply following the
instructions to the letter.

Has someone else had this error? Is there something that is "understood"
in the guide, that i could be missing?

Any help would be greatly appreciated!

//Jonas Adler

_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120724/9594150d/attachment.htm 

From finalguy at gmail.com  Wed Jul 25 03:26:56 2012
From: finalguy at gmail.com (Ming Ji)
Date: Tue, 24 Jul 2012 20:26:56 -0500
Subject: [zeromq-dev]  Question about constness
References: 500E28F6.1040608@gmail.com
Message-ID: <500F4B60.9030200@gmail.com>

Hi Sebastian,

Thanks for pointing out the previous threads about this issue. I went 
through all three of them. I saw someone mentioned const_cast. I would 
say that is a good way for people who knows the implementation details. 
For general users, it is still not safe. For example, zmq_msg_copy will 
destroy the original message. It is probably not a good idea to cast the 
const message and pass it to this function. We don't know if other 
functions will change the original messages, so it is in general not a 
good idea to blindly do the const_cast.

Could you share your interface, or C++ API patch? I can probably do a 
test run before you release it.

Thanks

From sebastian.lauwers at gmail.com  Wed Jul 25 11:58:11 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Wed, 25 Jul 2012 11:58:11 +0200
Subject: [zeromq-dev] Question about constness
In-Reply-To: <500F4B60.9030200@gmail.com>
References: 500E28F6.1040608@gmail.com <500F4B60.9030200@gmail.com>
Message-ID: <500FC333.1040307@gmail.com>

Hi again,

On 25/07/12 03:26, Ming Ji wrote:
> Could you share your interface, or C++ API patch? I can probably do a
> test run before you release it.

I haven't written the C++ API patch yet, so that'll have to wait for the 
weekend or so, when I have a minute. I've attached two files that 
contain my so called "interface".

Please note that this is a first try / prototype, so it will need to be 
modified quite a bit (very few socket options are supported) in order to 
respect everything ZeroQ has to offer, and to actually be usable in a 
generic way. I'm only using the minimum needed for my project, and will 
add to it when I reach the need for more features. This interface also 
assumes the use of Google's protobufs; but it would be easy to adapt to 
any other kind of serialisation library.

The way I use this in my project is as follows, components that "offer" 
a service, implement their "own socket":

class InstanceManagerSocket {
     private:
     firestarter::sockets::ZMQPublisherSocket publisher;
     firestarter::sockets::ZMQResponseSocket responder;

     public:
     InstanceManagerSocket(zmq::context_t & context) :
             publisher(context, MODULE_ORDERS_SOCKET_URI),
             responder(context, MANAGER_SOCKET_URI) { };
     inline bool send(google::protobuf::Message const & pb_message) {
         return this->publisher.send(pb_message);
     };
     inline bool reply(google::protobuf::Message const & pb_message) {
         return this->responder.send(pb_message);
     };
     inline bool receive(google::protobuf::Message & pb_message,
             bool blocking = false) {
         if (this->responder.receive(pb_message, blocking))
             return this->ack();
         return false;
      };
     inline bool ack() { return this->responder.send(); };
};

Clients who then want to use this service use a client class, for 
example (usually provided by the service):

class InstanceManagerClientSocket {
     private:
     firestarter::sockets::ZMQSubscriberSocket subscriber;
     firestarter::sockets::ZMQRequestSocket requester;

     public:
     InstanceManagerClientSocket(zmq::context_t & context) :
             subscriber(context, MODULE_ORDERS_SOCKET_URI),
             requester(context, MANAGER_SOCKET_URI) { };
     inline bool send(google::protobuf::Message const & pb_message) {
         if (this->requester.send(pb_message))
             return this->receive_ack();
         return false;
     };
     inline bool receive(google::protobuf::Message & pb_message,
             bool blocking = false) {
         return this->subscriber.receive(pb_message, blocking);
     };
     inline bool receive_ack() { return this->requester.receive(true); };
};


Again, please do take into account that this is in the very early stages 
of prototyping, and will probably be modified a lot before getting my 
own stamp of approval. What I'm aiming for, in the end, is to fully 
abstract the use of Protobufs for serialisation from the perspective of 
any one using the, for example, "InstanceManager sockets".

Regarding the actual ZeroMQ part, I'll probably change the design of the 
classes, most probably through the use of templates, in order to remove 
the small artifacts that exist today (connect() or bind() end up in the 
wrong sockets, due to the use of virtual multiple inheritance, which I'm 
not happy with).

@Everyone: I welcome any comments regarding my current use of 
ZeroMQ---my understanding of it is basic to say the least, which is why 
I haven't attacked more consuming bits like devices and such.

-S.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: zmqhelper.h
Type: text/x-chdr
Size: 196 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/bc2dbe13/attachment.h 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: zmqsocket.cpp
Type: text/x-c++src
Size: 2269 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/bc2dbe13/attachment.cpp 

From Sharon.Ben-Asher at avg.com  Wed Jul 25 12:14:03 2012
From: Sharon.Ben-Asher at avg.com (Sharon Ben-Asher)
Date: Wed, 25 Jul 2012 10:14:03 +0000
Subject: [zeromq-dev] zeromq-dev Digest, Vol 55, Issue 23
In-Reply-To: <mailman.1.1343210402.2715.zeromq-dev@lists.zeromq.org>
References: <mailman.1.1343210402.2715.zeromq-dev@lists.zeromq.org>
Message-ID: <671E27F33F2F25499F78A4AFA494E63B384D8010@ex10czdmb01>

Jonas,

I have another idea (in case you did make zmq before jzmq)
Notice that the configure of both zmq and jzmq rely on correct setting of $OUTPUT_DIR env variable. 
Specifically, you can see this in the configure of jzmq 
LDFLAGS="-L$OUTPUT_DIR/lib"
This tells the linker of jzmq where to find libzmq.so
So make sure the env variable is set correctly.

Good luck,

Sharon.


-----Original Message-----
From: jonas.adler at epiq.se
Subject: [zeromq-dev] Massive problems running ZMQ on android
To: zeromq-dev at lists.zeromq.org
Message-ID: <a469afc041f1b0e985d760584d6d0866.squirrel at mail.glesys.se>
Content-Type: text/plain;charset=iso-8859-1

Hello,

I'm new to ZMQ, and trying to get it to run on android. I am however, facing issue, after issue, after issue.

I try to follow the official guide, (http://www.zeromq.org/build:android),
and have gotten to the following call on my ubuntu machine:

./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version

I get this error:

"configure: error: cannot link with -lzmq"

I have tried google for a few days now, but i cannot find any solution to this that works for me.

Some comments on other solutions i have found, that did not work for me:

I have installed pretty much every compiler known to man, but if you have any specific suggestions, please tell me.
I have made no alterations to the source at all, simply following the instructions to the letter.

Has someone else had this error? Is there something that is "understood"
in the guide, that i could be missing?

Any help would be greatly appreciated!

//Jonas Adler


From jonas.adler at epiq.se  Wed Jul 25 13:30:40 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Wed, 25 Jul 2012 13:30:40 +0200 (CEST)
Subject: [zeromq-dev] Re :  Massive problems running ZMQ on android
In-Reply-To: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
Message-ID: <df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>

Yes, I configured zmq before I tried to configure jzmq.
My OUTPUT_DIR was set to /tmp/zeromq-android before I configured zmq.

Here's a link to the script I made (from the instructions on
build:android) after redoing the instructions approximate 15 times.

Worth noting is that I installed the NDK and added the
/opt/android-toolchain/bin to my PATH by hand, the steps i followed are at
the top of this pastebin post.

http://pastebin.com/iXQWCxNV

Everything seems to go smooth untill the jzmq config. That is where i get
the "configure: error: cannot link with -lzmq" error.

Everything I did tried I did with sudo or root access.

> Did you try to configure zmq or jzmq?The configure command you described
> is command line for jzmq. You need first to have set the env
> variable?OUTPUT_DIR=/tmp/zeromq-android?and then have generated zmq.
> Then you should be able to launch configure for jzmq.
> Regards.
>
> --- En date de?: Mar 24.7.12, jonas.adler at epiq.se <jonas.adler at epiq.se> a
> ?crit?:
>
> De: jonas.adler at epiq.se <jonas.adler at epiq.se>
> Objet: [zeromq-dev] Massive problems running ZMQ on android
> ?: zeromq-dev at lists.zeromq.org
> Date: Mardi 24 juillet 2012, 16h03
>
> Hello,
>
> I'm new to ZMQ, and trying to get it to run on android. I am however,
> facing issue, after issue, after issue.
>
> I try to follow the official guide, (http://www.zeromq.org/build:android),
> and have gotten to the following call on my ubuntu machine:
>
> ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
>
> I get this error:
>
> "configure: error: cannot link with -lzmq"
>
> I have tried google for a few days now, but i cannot find any solution to
> this that works for me.
>
> Some comments on other solutions i have found, that did not work for me:
>
> I have installed pretty much every compiler known to man, but if you have
> any specific suggestions, please tell me.
> I have made no alterations to the source at all, simply following the
> instructions to the letter.
>
> Has someone else had this error? Is there something that is "understood"
> in the guide, that i could be missing?
>
> Any help would be greatly appreciated!
>
> //Jonas Adler
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>


From victor at iso3103.net  Wed Jul 25 14:05:08 2012
From: victor at iso3103.net (Victor Perron)
Date: Wed, 25 Jul 2012 14:05:08 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
Message-ID: <CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>

OK, I ran the exact same sequence. Seems you are right, I have forgotten a
little word on the last configure line.
When it fails, the config.log says:

504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function `uuid_t':
505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
`uuid_generate'
506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to `uuid_unparse'
507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
`uuid_generate'
508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to `uuid_unparse'

Well if you recompile jzmq with the line:

./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
--with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version

and everything seems fine. Can you confirm it works ?

Best regards,

On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:

> Yes, I configured zmq before I tried to configure jzmq.
> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured zmq.
>
> Here's a link to the script I made (from the instructions on
> build:android) after redoing the instructions approximate 15 times.
>
> Worth noting is that I installed the NDK and added the
> /opt/android-toolchain/bin to my PATH by hand, the steps i followed are at
> the top of this pastebin post.
>
> http://pastebin.com/iXQWCxNV
>
> Everything seems to go smooth untill the jzmq config. That is where i get
> the "configure: error: cannot link with -lzmq" error.
>
> Everything I did tried I did with sudo or root access.
>
> > Did you try to configure zmq or jzmq?The configure command you described
> > is command line for jzmq. You need first to have set the env
> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated zmq.
> > Then you should be able to launch configure for jzmq.
> > Regards.
> >
> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se <jonas.adler at epiq.se>
> a
> > ?crit :
> >
> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
> > Objet: [zeromq-dev] Massive problems running ZMQ on android
> > ?: zeromq-dev at lists.zeromq.org
> > Date: Mardi 24 juillet 2012, 16h03
> >
> > Hello,
> >
> > I'm new to ZMQ, and trying to get it to run on android. I am however,
> > facing issue, after issue, after issue.
> >
> > I try to follow the official guide, (http://www.zeromq.org/build:android
> ),
> > and have gotten to the following call on my ubuntu machine:
> >
> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
> >
> > I get this error:
> >
> > "configure: error: cannot link with -lzmq"
> >
> > I have tried google for a few days now, but i cannot find any solution to
> > this that works for me.
> >
> > Some comments on other solutions i have found, that did not work for me:
> >
> > I have installed pretty much every compiler known to man, but if you have
> > any specific suggestions, please tell me.
> > I have made no alterations to the source at all, simply following the
> > instructions to the letter.
> >
> > Has someone else had this error? Is there something that is "understood"
> > in the guide, that i could be missing?
> >
> > Any help would be greatly appreciated!
> >
> > //Jonas Adler
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/6d99ac46/attachment.htm 

From jonas.adler at epiq.se  Wed Jul 25 15:17:04 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Wed, 25 Jul 2012 15:17:04 +0200 (CEST)
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
Message-ID: <29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>

I still get the same problem with zmq_init not in -lzmq.

some enviroment variables that might be useful:

PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games

OUTPUT_DIR=/tmp/zeromq-android

packages I installed (running ubuntu x86_64 clean install before i tried)
was:
 * libtool, autogen, automake
 * gcc, gpp, gcc-multilib
 * libc6-dev
 * git

something I'm missing?

The tarball for zmq I downloaded was
http://download.zeromq.org/zeromq-2.2.0.tar.gz
(with wget into /tmp/. Since the wiki doesn't state where to get it i just
took it from the download page)

Also here is the config.log file the JZMQ ./config generated (with that
LIBS"-luuid" added to the flags, as you can see on line 7)

Thanks in advance.

> OK, I ran the exact same sequence. Seems you are right, I have forgotten a
> little word on the last configure line.
> When it fails, the config.log says:
>
> 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function `uuid_t':
> 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> `uuid_generate'
> 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> `uuid_unparse'
> 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> `uuid_generate'
> 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> `uuid_unparse'
>
> Well if you recompile jzmq with the line:
>
> ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
>
> and everything seems fine. Can you confirm it works ?
>
> Best regards,
>
> On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
>
>> Yes, I configured zmq before I tried to configure jzmq.
>> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured zmq.
>>
>> Here's a link to the script I made (from the instructions on
>> build:android) after redoing the instructions approximate 15 times.
>>
>> Worth noting is that I installed the NDK and added the
>> /opt/android-toolchain/bin to my PATH by hand, the steps i followed are
>> at
>> the top of this pastebin post.
>>
>> http://pastebin.com/iXQWCxNV
>>
>> Everything seems to go smooth untill the jzmq config. That is where i
>> get
>> the "configure: error: cannot link with -lzmq" error.
>>
>> Everything I did tried I did with sudo or root access.
>>
>> > Did you try to configure zmq or jzmq?The configure command you
>> described
>> > is command line for jzmq. You need first to have set the env
>> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated zmq.
>> > Then you should be able to launch configure for jzmq.
>> > Regards.
>> >
>> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
>> <jonas.adler at epiq.se>
>> a
>> > ?crit :
>> >
>> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
>> > Objet: [zeromq-dev] Massive problems running ZMQ on android
>> > ?: zeromq-dev at lists.zeromq.org
>> > Date: Mardi 24 juillet 2012, 16h03
>> >
>> > Hello,
>> >
>> > I'm new to ZMQ, and trying to get it to run on android. I am however,
>> > facing issue, after issue, after issue.
>> >
>> > I try to follow the official guide,
>> (http://www.zeromq.org/build:android
>> ),
>> > and have gotten to the following call on my ubuntu machine:
>> >
>> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
>> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
>> >
>> > I get this error:
>> >
>> > "configure: error: cannot link with -lzmq"
>> >
>> > I have tried google for a few days now, but i cannot find any solution
>> to
>> > this that works for me.
>> >
>> > Some comments on other solutions i have found, that did not work for
>> me:
>> >
>> > I have installed pretty much every compiler known to man, but if you
>> have
>> > any specific suggestions, please tell me.
>> > I have made no alterations to the source at all, simply following the
>> > instructions to the letter.
>> >
>> > Has someone else had this error? Is there something that is
>> "understood"
>> > in the guide, that i could be missing?
>> >
>> > Any help would be greatly appreciated!
>> >
>> > //Jonas Adler
>> >
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>>
>>
>
>
> --
> Victor
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>


From victor at iso3103.net  Wed Jul 25 16:13:28 2012
From: victor at iso3103.net (Victor Perron)
Date: Wed, 25 Jul 2012 16:13:28 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
Message-ID: <CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>

Cqn you please send me your config.log ?

BR,

On Wed, Jul 25, 2012 at 3:17 PM, <jonas.adler at epiq.se> wrote:

> I still get the same problem with zmq_init not in -lzmq.
>
> some enviroment variables that might be useful:
>
>
> PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games
>
> OUTPUT_DIR=/tmp/zeromq-android
>
> packages I installed (running ubuntu x86_64 clean install before i tried)
> was:
>  * libtool, autogen, automake
>  * gcc, gpp, gcc-multilib
>  * libc6-dev
>  * git
>
> something I'm missing?
>
> The tarball for zmq I downloaded was
> http://download.zeromq.org/zeromq-2.2.0.tar.gz
> (with wget into /tmp/. Since the wiki doesn't state where to get it i just
> took it from the download page)
>
> Also here is the config.log file the JZMQ ./config generated (with that
> LIBS"-luuid" added to the flags, as you can see on line 7)
>
> Thanks in advance.
>
> > OK, I ran the exact same sequence. Seems you are right, I have forgotten
> a
> > little word on the last configure line.
> > When it fails, the config.log says:
> >
> > 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function `uuid_t':
> > 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> > `uuid_generate'
> > 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> > `uuid_unparse'
> > 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> > `uuid_generate'
> > 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> > `uuid_unparse'
> >
> > Well if you recompile jzmq with the line:
> >
> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> > LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
> >
> > and everything seems fine. Can you confirm it works ?
> >
> > Best regards,
> >
> > On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
> >
> >> Yes, I configured zmq before I tried to configure jzmq.
> >> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured zmq.
> >>
> >> Here's a link to the script I made (from the instructions on
> >> build:android) after redoing the instructions approximate 15 times.
> >>
> >> Worth noting is that I installed the NDK and added the
> >> /opt/android-toolchain/bin to my PATH by hand, the steps i followed are
> >> at
> >> the top of this pastebin post.
> >>
> >> http://pastebin.com/iXQWCxNV
> >>
> >> Everything seems to go smooth untill the jzmq config. That is where i
> >> get
> >> the "configure: error: cannot link with -lzmq" error.
> >>
> >> Everything I did tried I did with sudo or root access.
> >>
> >> > Did you try to configure zmq or jzmq?The configure command you
> >> described
> >> > is command line for jzmq. You need first to have set the env
> >> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated zmq.
> >> > Then you should be able to launch configure for jzmq.
> >> > Regards.
> >> >
> >> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
> >> <jonas.adler at epiq.se>
> >> a
> >> > ?crit :
> >> >
> >> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
> >> > Objet: [zeromq-dev] Massive problems running ZMQ on android
> >> > ?: zeromq-dev at lists.zeromq.org
> >> > Date: Mardi 24 juillet 2012, 16h03
> >> >
> >> > Hello,
> >> >
> >> > I'm new to ZMQ, and trying to get it to run on android. I am however,
> >> > facing issue, after issue, after issue.
> >> >
> >> > I try to follow the official guide,
> >> (http://www.zeromq.org/build:android
> >> ),
> >> > and have gotten to the following call on my ubuntu machine:
> >> >
> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> >> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
> >> >
> >> > I get this error:
> >> >
> >> > "configure: error: cannot link with -lzmq"
> >> >
> >> > I have tried google for a few days now, but i cannot find any solution
> >> to
> >> > this that works for me.
> >> >
> >> > Some comments on other solutions i have found, that did not work for
> >> me:
> >> >
> >> > I have installed pretty much every compiler known to man, but if you
> >> have
> >> > any specific suggestions, please tell me.
> >> > I have made no alterations to the source at all, simply following the
> >> > instructions to the letter.
> >> >
> >> > Has someone else had this error? Is there something that is
> >> "understood"
> >> > in the guide, that i could be missing?
> >> >
> >> > Any help would be greatly appreciated!
> >> >
> >> > //Jonas Adler
> >> >
> >> > _______________________________________________
> >> > zeromq-dev mailing list
> >> > zeromq-dev at lists.zeromq.org
> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> > _______________________________________________
> >> > zeromq-dev mailing list
> >> > zeromq-dev at lists.zeromq.org
> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >
> >>
> >>
> >
> >
> > --
> > Victor
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/ccde4784/attachment.htm 

From jonas.adler at epiq.se  Wed Jul 25 16:20:02 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Wed, 25 Jul 2012 16:20:02 +0200 (CEST)
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
Message-ID: <e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>

Oh i forgot to add the link to the log, I'm sorry :p

http://pastebin.com/pWwepPdr

> Cqn you please send me your config.log ?
>
> BR,
>
> On Wed, Jul 25, 2012 at 3:17 PM, <jonas.adler at epiq.se> wrote:
>
>> I still get the same problem with zmq_init not in -lzmq.
>>
>> some enviroment variables that might be useful:
>>
>>
>> PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games
>>
>> OUTPUT_DIR=/tmp/zeromq-android
>>
>> packages I installed (running ubuntu x86_64 clean install before i
>> tried)
>> was:
>>  * libtool, autogen, automake
>>  * gcc, gpp, gcc-multilib
>>  * libc6-dev
>>  * git
>>
>> something I'm missing?
>>
>> The tarball for zmq I downloaded was
>> http://download.zeromq.org/zeromq-2.2.0.tar.gz
>> (with wget into /tmp/. Since the wiki doesn't state where to get it i
>> just
>> took it from the download page)
>>
>> Also here is the config.log file the JZMQ ./config generated (with that
>> LIBS"-luuid" added to the flags, as you can see on line 7)
>>
>> Thanks in advance.
>>
>> > OK, I ran the exact same sequence. Seems you are right, I have
>> forgotten
>> a
>> > little word on the last configure line.
>> > When it fails, the config.log says:
>> >
>> > 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function
>> `uuid_t':
>> > 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
>> > `uuid_generate'
>> > 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
>> > `uuid_unparse'
>> > 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
>> > `uuid_generate'
>> > 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
>> > `uuid_unparse'
>> >
>> > Well if you recompile jzmq with the line:
>> >
>> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
>> > LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
>> >
>> > and everything seems fine. Can you confirm it works ?
>> >
>> > Best regards,
>> >
>> > On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
>> >
>> >> Yes, I configured zmq before I tried to configure jzmq.
>> >> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured zmq.
>> >>
>> >> Here's a link to the script I made (from the instructions on
>> >> build:android) after redoing the instructions approximate 15 times.
>> >>
>> >> Worth noting is that I installed the NDK and added the
>> >> /opt/android-toolchain/bin to my PATH by hand, the steps i followed
>> are
>> >> at
>> >> the top of this pastebin post.
>> >>
>> >> http://pastebin.com/iXQWCxNV
>> >>
>> >> Everything seems to go smooth untill the jzmq config. That is where i
>> >> get
>> >> the "configure: error: cannot link with -lzmq" error.
>> >>
>> >> Everything I did tried I did with sudo or root access.
>> >>
>> >> > Did you try to configure zmq or jzmq?The configure command you
>> >> described
>> >> > is command line for jzmq. You need first to have set the env
>> >> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated
>> zmq.
>> >> > Then you should be able to launch configure for jzmq.
>> >> > Regards.
>> >> >
>> >> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
>> >> <jonas.adler at epiq.se>
>> >> a
>> >> > ?crit :
>> >> >
>> >> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
>> >> > Objet: [zeromq-dev] Massive problems running ZMQ on android
>> >> > ?: zeromq-dev at lists.zeromq.org
>> >> > Date: Mardi 24 juillet 2012, 16h03
>> >> >
>> >> > Hello,
>> >> >
>> >> > I'm new to ZMQ, and trying to get it to run on android. I am
>> however,
>> >> > facing issue, after issue, after issue.
>> >> >
>> >> > I try to follow the official guide,
>> >> (http://www.zeromq.org/build:android
>> >> ),
>> >> > and have gotten to the following call on my ubuntu machine:
>> >> >
>> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
>> >> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
>> >> >
>> >> > I get this error:
>> >> >
>> >> > "configure: error: cannot link with -lzmq"
>> >> >
>> >> > I have tried google for a few days now, but i cannot find any
>> solution
>> >> to
>> >> > this that works for me.
>> >> >
>> >> > Some comments on other solutions i have found, that did not work
>> for
>> >> me:
>> >> >
>> >> > I have installed pretty much every compiler known to man, but if
>> you
>> >> have
>> >> > any specific suggestions, please tell me.
>> >> > I have made no alterations to the source at all, simply following
>> the
>> >> > instructions to the letter.
>> >> >
>> >> > Has someone else had this error? Is there something that is
>> >> "understood"
>> >> > in the guide, that i could be missing?
>> >> >
>> >> > Any help would be greatly appreciated!
>> >> >
>> >> > //Jonas Adler
>> >> >
>> >> > _______________________________________________
>> >> > zeromq-dev mailing list
>> >> > zeromq-dev at lists.zeromq.org
>> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> > _______________________________________________
>> >> > zeromq-dev mailing list
>> >> > zeromq-dev at lists.zeromq.org
>> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >
>> >>
>> >>
>> >
>> >
>> > --
>> > Victor
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>>
>>
>
>
> --
> Victor
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



From victor at iso3103.net  Wed Jul 25 16:52:18 2012
From: victor at iso3103.net (Victor Perron)
Date: Wed, 25 Jul 2012 16:52:18 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
Message-ID: <CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>

Seems there is an issue in your PATH. From the line 83 (and the others) it
looks like the toolchain is not found:

configure:3246: checking for arm-linux-androideabi-gcc
configure:3276: result: no
configure:3286: checking for gcc
configure:3302: found /usr/bin/gcc
configure:3313: result: gcc

... and so on. In the end the compiler complains about clock_gettime() and
pthread_create not found, but that;s an aftereffect of not using the right
compiler.

Could you check your $PATH ?

On Wed, Jul 25, 2012 at 4:20 PM, <jonas.adler at epiq.se> wrote:

> Oh i forgot to add the link to the log, I'm sorry :p
>
> http://pastebin.com/pWwepPdr
>
> > Cqn you please send me your config.log ?
> >
> > BR,
> >
> > On Wed, Jul 25, 2012 at 3:17 PM, <jonas.adler at epiq.se> wrote:
> >
> >> I still get the same problem with zmq_init not in -lzmq.
> >>
> >> some enviroment variables that might be useful:
> >>
> >>
> >>
> PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games
> >>
> >> OUTPUT_DIR=/tmp/zeromq-android
> >>
> >> packages I installed (running ubuntu x86_64 clean install before i
> >> tried)
> >> was:
> >>  * libtool, autogen, automake
> >>  * gcc, gpp, gcc-multilib
> >>  * libc6-dev
> >>  * git
> >>
> >> something I'm missing?
> >>
> >> The tarball for zmq I downloaded was
> >> http://download.zeromq.org/zeromq-2.2.0.tar.gz
> >> (with wget into /tmp/. Since the wiki doesn't state where to get it i
> >> just
> >> took it from the download page)
> >>
> >> Also here is the config.log file the JZMQ ./config generated (with that
> >> LIBS"-luuid" added to the flags, as you can see on line 7)
> >>
> >> Thanks in advance.
> >>
> >> > OK, I ran the exact same sequence. Seems you are right, I have
> >> forgotten
> >> a
> >> > little word on the last configure line.
> >> > When it fails, the config.log says:
> >> >
> >> > 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function
> >> `uuid_t':
> >> > 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> >> > `uuid_generate'
> >> > 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> >> > `uuid_unparse'
> >> > 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> >> > `uuid_generate'
> >> > 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> >> > `uuid_unparse'
> >> >
> >> > Well if you recompile jzmq with the line:
> >> >
> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> >> > LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
> >> >
> >> > and everything seems fine. Can you confirm it works ?
> >> >
> >> > Best regards,
> >> >
> >> > On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
> >> >
> >> >> Yes, I configured zmq before I tried to configure jzmq.
> >> >> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured zmq.
> >> >>
> >> >> Here's a link to the script I made (from the instructions on
> >> >> build:android) after redoing the instructions approximate 15 times.
> >> >>
> >> >> Worth noting is that I installed the NDK and added the
> >> >> /opt/android-toolchain/bin to my PATH by hand, the steps i followed
> >> are
> >> >> at
> >> >> the top of this pastebin post.
> >> >>
> >> >> http://pastebin.com/iXQWCxNV
> >> >>
> >> >> Everything seems to go smooth untill the jzmq config. That is where i
> >> >> get
> >> >> the "configure: error: cannot link with -lzmq" error.
> >> >>
> >> >> Everything I did tried I did with sudo or root access.
> >> >>
> >> >> > Did you try to configure zmq or jzmq?The configure command you
> >> >> described
> >> >> > is command line for jzmq. You need first to have set the env
> >> >> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated
> >> zmq.
> >> >> > Then you should be able to launch configure for jzmq.
> >> >> > Regards.
> >> >> >
> >> >> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
> >> >> <jonas.adler at epiq.se>
> >> >> a
> >> >> > ?crit :
> >> >> >
> >> >> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
> >> >> > Objet: [zeromq-dev] Massive problems running ZMQ on android
> >> >> > ?: zeromq-dev at lists.zeromq.org
> >> >> > Date: Mardi 24 juillet 2012, 16h03
> >> >> >
> >> >> > Hello,
> >> >> >
> >> >> > I'm new to ZMQ, and trying to get it to run on android. I am
> >> however,
> >> >> > facing issue, after issue, after issue.
> >> >> >
> >> >> > I try to follow the official guide,
> >> >> (http://www.zeromq.org/build:android
> >> >> ),
> >> >> > and have gotten to the following call on my ubuntu machine:
> >> >> >
> >> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> >> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> >> >> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
> >> >> >
> >> >> > I get this error:
> >> >> >
> >> >> > "configure: error: cannot link with -lzmq"
> >> >> >
> >> >> > I have tried google for a few days now, but i cannot find any
> >> solution
> >> >> to
> >> >> > this that works for me.
> >> >> >
> >> >> > Some comments on other solutions i have found, that did not work
> >> for
> >> >> me:
> >> >> >
> >> >> > I have installed pretty much every compiler known to man, but if
> >> you
> >> >> have
> >> >> > any specific suggestions, please tell me.
> >> >> > I have made no alterations to the source at all, simply following
> >> the
> >> >> > instructions to the letter.
> >> >> >
> >> >> > Has someone else had this error? Is there something that is
> >> >> "understood"
> >> >> > in the guide, that i could be missing?
> >> >> >
> >> >> > Any help would be greatly appreciated!
> >> >> >
> >> >> > //Jonas Adler
> >> >> >
> >> >> > _______________________________________________
> >> >> > zeromq-dev mailing list
> >> >> > zeromq-dev at lists.zeromq.org
> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >> > _______________________________________________
> >> >> > zeromq-dev mailing list
> >> >> > zeromq-dev at lists.zeromq.org
> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >> >
> >> >>
> >> >>
> >> >
> >> >
> >> > --
> >> > Victor
> >> > _______________________________________________
> >> > zeromq-dev mailing list
> >> > zeromq-dev at lists.zeromq.org
> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >
> >>
> >>
> >
> >
> > --
> > Victor
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
>
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/c3e3ffc5/attachment.htm 

From mrossi19 at gmail.com  Wed Jul 25 18:11:50 2012
From: mrossi19 at gmail.com (Marc Rossi)
Date: Wed, 25 Jul 2012 11:11:50 -0500
Subject: [zeromq-dev] Connect with inproc
Message-ID: <CAMJu8UZz4_acfHfDw+YhdoCjMSAbNO=OoBJvNudtAxEw1_6sbQ@mail.gmail.com>

Linux w/ ZMQ 2.2.0

I have a publish function that is called from a thread.  This publish
function connects an inproc socket with a socket in another thread.
Since inproc sockets error on a connect call if the other side isn't
setup yet I have the publish functino return without publishing
until the connection can be succesfully established.

This function publishes a high volume of data and sometimes when the
other side of the socket takes a second or two to setup I run into
the assertion

    Too many open files (signaler.cpp:330)

when it tries to create a socketpair.  Can't find where the socketpair
gets cleaned up after deletion.

Am I missing something with my approach? Initially tried to reuse
created socket instead of deleting/recreating every time but
connect would never succeed.

TIA
Marc

zmq::socket *s = NULL;
void publish(void *data)
{
    if (s == NULL)
    {
        s = new zmq::socket_t(context, ZMQ_PUB);

        try
        {
            s->connect("inproc://test");
        }

        catch (...)
        {
            delete s;
            s = NULL;
            return;
        }
    }

    // here I publish data on the socket
}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/fcc4e494/attachment.htm 

From lists at chuckremes.com  Wed Jul 25 18:45:11 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Wed, 25 Jul 2012 11:45:11 -0500
Subject: [zeromq-dev] Connect with inproc
In-Reply-To: <CAMJu8UZz4_acfHfDw+YhdoCjMSAbNO=OoBJvNudtAxEw1_6sbQ@mail.gmail.com>
References: <CAMJu8UZz4_acfHfDw+YhdoCjMSAbNO=OoBJvNudtAxEw1_6sbQ@mail.gmail.com>
Message-ID: <9852F3B7-D494-4CF6-A996-5DE6D6BEBEB7@chuckremes.com>


On Jul 25, 2012, at 11:11 AM, Marc Rossi wrote:

> 
> Linux w/ ZMQ 2.2.0
> 
> I have a publish function that is called from a thread.  This publish
> function connects an inproc socket with a socket in another thread.
> Since inproc sockets error on a connect call if the other side isn't
> setup yet I have the publish functino return without publishing
> until the connection can be succesfully established.
> 
> This function publishes a high volume of data and sometimes when the
> other side of the socket takes a second or two to setup I run into
> the assertion
> 
>     Too many open files (signaler.cpp:330)
> 
> when it tries to create a socketpair.  Can't find where the socketpair
> gets cleaned up after deletion.
> 
> Am I missing something with my approach? Initially tried to reuse
> created socket instead of deleting/recreating every time but
> connect would never succeed.
> 

Unless there is some delay between attempts, you are probably trying to create and connect that socket hundreds or thousands of times per second until the other thread has completed its setup and called zmq_bind() on the inproc transport.

Socket deletion, like everything else in zeromq, is asynchronous. So, while you may have deleted the socket it does not mean that its resources have been reclaimed yet. In a tight enough loop you will run out of file descriptors because the "deleted" sockets are still holding on to unreleased file descriptors while they await their termination.

I suggest using a real signaling mechanism (like a condition variable) to indicate that your receiver thread has completed its setup. Only then should you try and create the publish socket and connect it.

cr


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120725/5908cd4a/attachment.htm 

From Richard_Newton at waters.com  Thu Jul 26 14:31:29 2012
From: Richard_Newton at waters.com (Richard_Newton at waters.com)
Date: Thu, 26 Jul 2012 13:31:29 +0100
Subject: [zeromq-dev] [PATCH] [cppzmq] Add move constructor to message_t
Message-ID: <OF21D5BAA1.FE45B344-ON80257A47.00446023-80257A47.0044CCF8@waters.com>


Hi,

In the C++ binding both context_t and socket_t have move constructors, I would like to extend that to message_t as well.


(See attached file: Add-move-constructor-to-message_t.patch)

This was again the master branch on github, also sent corresponding pull request.

Regards,

Richard.
===========================================================
The information in this email is confidential, and is intended solely for the addressee(s). 
Access to this email by anyone else is unauthorized and therefore prohibited.  If you are 
not the intended recipient you are notified that disclosing, copying, distributing or taking 
any action in reliance on the contents of this information is strictly prohibited and may be unlawful.
===========================================================
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120726/361306d0/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Add-move-constructor-to-message_t.patch
Type: application/octet-stream
Size: 1034 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120726/361306d0/attachment.obj 

From linkfanel at yahoo.fr  Thu Jul 26 17:27:30 2012
From: linkfanel at yahoo.fr (Pierre Ynard)
Date: Thu, 26 Jul 2012 17:27:30 +0200
Subject: [zeromq-dev] ZMQ_MCAST_LOOP with PGM
Message-ID: <20120726152730.GA16824@via.ecp.fr>

Hello,

I took notice of https://zeromq.jira.com/browse/LIBZMQ-104
and of the removal of this option, but I have
trouble understanding the issue. Even after reading
http://lists.zeromq.org/pipermail/zeromq-dev/2010-June/004133.html
I don't get it.

> unicast packets are not broadcast to all listening sockets hence
> back channel communication is only functional to the first opened
> transport.

Are we talking sender or receiver side of the transport here? What is
specific to loopback with this? Isn't it impossible to have several
sockets listening on the same unicast address anyway?

Anyway, is it safe to enable this option if I don't mind about
reliability?

-- 
Pierre Ynard
"Une ?me dans un corps, c'est comme un dessin sur une feuille de papier."

From ian.barber at gmail.com  Thu Jul 26 17:35:10 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Thu, 26 Jul 2012 16:35:10 +0100
Subject: [zeromq-dev] ZMQ_MCAST_LOOP with PGM
In-Reply-To: <20120726152730.GA16824@via.ecp.fr>
References: <20120726152730.GA16824@via.ecp.fr>
Message-ID: <CAEYnEwEv31kATxJsOA=-Ja87VwV-bQpV5w4RZO7FoeKjwNYjjA@mail.gmail.com>

On Thu, Jul 26, 2012 at 4:27 PM, Pierre Ynard <linkfanel at yahoo.fr> wrote:

>
> Anyway, is it safe to enable this option if I don't mind about
> reliability?
>

I'm sure Steve can give a better answer, but basically multicast
loopback just doesn't work very well at all, and causes a lot of
confusion among people trying out PGM.

Ian

From steven.mccoy at miru.hk  Thu Jul 26 20:27:15 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Thu, 26 Jul 2012 14:27:15 -0400
Subject: [zeromq-dev] ZMQ_MCAST_LOOP with PGM
In-Reply-To: <CAEYnEwEv31kATxJsOA=-Ja87VwV-bQpV5w4RZO7FoeKjwNYjjA@mail.gmail.com>
References: <20120726152730.GA16824@via.ecp.fr>
	<CAEYnEwEv31kATxJsOA=-Ja87VwV-bQpV5w4RZO7FoeKjwNYjjA@mail.gmail.com>
Message-ID: <CAGFWLigj7NWGic6RgNfSkWLpZuhAM60E1rjgsCpLrKBCQuZU4w@mail.gmail.com>

On 26 July 2012 11:35, Ian Barber <ian.barber at gmail.com> wrote:

> On Thu, Jul 26, 2012 at 4:27 PM, Pierre Ynard <linkfanel at yahoo.fr> wrote:
>
> >
> > Anyway, is it safe to enable this option if I don't mind about
> > reliability?
> >
>
> I'm sure Steve can give a better answer, but basically multicast
> loopback just doesn't work very well at all, and causes a lot of
> confusion among people trying out PGM.
>
>
It can work well in a restricted environment, the disruptor team have shown
great performance with Java and multicast loop but with the PGM protocol
specifically there are problems and limitations that can break reliability.

From steven.mccoy at miru.hk  Fri Jul 27 03:54:10 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Thu, 26 Jul 2012 21:54:10 -0400
Subject: [zeromq-dev] Build errors with Win64 on CZMQ
Message-ID: <CAGFWLigONd1e3q2KcKwhCUGy6Eh223w-AwfEyijd5x_VQ2GUaw@mail.gmail.com>

Looks like the code is not both C99 and C++2003 compliant:

C:\temp\czmq-1.1.0\src\zhash.c(131) : warning C4996: 'strdup': The POSIX
name for this item is deprecated. Instead, use the ISO C++ conformant name:
_strdup. See online help for details.
        c:\Program Files (x86)\Microsoft Visual Studio
10.0\VC\INCLUDE\string.h(238) : see declaration of 'strdup'
C:\temp\czmq-1.1.0\src\zmsg.c(179) : error C2440: 'initializing' : cannot
convert from 'void *' to 'zframe_t *'
        Conversion from 'void*' to pointer to non-'void' requires an
explicit cast
C:\temp\czmq-1.1.0\src\zmsg.c(462) : error C2440: '=' : cannot convert from
'void *' to 'byte *'
        Conversion from 'void*' to pointer to non-'void' requires an
explicit cast
C:\temp\czmq-1.1.0\src\zmsg.c(680) : error C2440: 'initializing' : cannot
convert from 'void *' to 'byte *'
        Conversion from 'void*' to pointer to non-'void' requires an
explicit cast

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120726/5459d08c/attachment.htm 

From steven.mccoy at miru.hk  Fri Jul 27 04:38:27 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Thu, 26 Jul 2012 22:38:27 -0400
Subject: [zeromq-dev] Build errors with Win64 on CZMQ
In-Reply-To: <CAGFWLigONd1e3q2KcKwhCUGy6Eh223w-AwfEyijd5x_VQ2GUaw@mail.gmail.com>
References: <CAGFWLigONd1e3q2KcKwhCUGy6Eh223w-AwfEyijd5x_VQ2GUaw@mail.gmail.com>
Message-ID: <CAGFWLig9QR0WgtfxhiLQQshG2gA3qStJhKh=focsVqxWmfNctQ@mail.gmail.com>

I found a ssize_t in there too, Windows only has SSIZE_T.

-- 
Steve-o

On 26 July 2012 21:54, Steven McCoy <steven.mccoy at miru.hk> wrote:

> Looks like the code is not both C99 and C++2003 compliant:
>
> C:\temp\czmq-1.1.0\src\zhash.c(131) : warning C4996: 'strdup': The POSIX
> name for this item is deprecated. Instead, use the ISO C++ conformant name:
> _strdup. See online help for details.
>         c:\Program Files (x86)\Microsoft Visual Studio
> 10.0\VC\INCLUDE\string.h(238) : see declaration of 'strdup'
> C:\temp\czmq-1.1.0\src\zmsg.c(179) : error C2440: 'initializing' : cannot
> convert from 'void *' to 'zframe_t *'
>         Conversion from 'void*' to pointer to non-'void' requires an
> explicit cast
> C:\temp\czmq-1.1.0\src\zmsg.c(462) : error C2440: '=' : cannot convert
> from 'void *' to 'byte *'
>         Conversion from 'void*' to pointer to non-'void' requires an
> explicit cast
> C:\temp\czmq-1.1.0\src\zmsg.c(680) : error C2440: 'initializing' : cannot
> convert from 'void *' to 'byte *'
>         Conversion from 'void*' to pointer to non-'void' requires an
> explicit cast
>
> --
> Steve-o
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120726/9c509365/attachment.htm 

From finalguy at gmail.com  Thu Jul 26 04:50:23 2012
From: finalguy at gmail.com (Ming Ji)
Date: Wed, 25 Jul 2012 21:50:23 -0500
Subject: [zeromq-dev]   Question about constness
Message-ID: <5010B06F.1080506@gmail.com>

I was wrong that zmq_msg_copy does not destroy the original message. It 
just does a shadow copy. Still it is better to indicate if the functions 
might change the original message at API level.

From ivan at blackpx.com  Thu Jul 26 21:29:53 2012
From: ivan at blackpx.com (Ivan Figueredo)
Date: Thu, 26 Jul 2012 14:29:53 -0500
Subject: [zeromq-dev] If not Majordomo for asynchronous workers, then what?
Message-ID: <CAEiCcQBXNrGyWqj17tgTDzJw4gVqmHp7Vj-QoJYj4zNYnaMgow@mail.gmail.com>

I am writing a distributed trading manager. There are many different
liquidity providers, and the client would like to route to a given
exchange. With this in mind, I thought that using the Majordomo pattern
with a broker in between clients and workers asking a particular [named]
worker to do its work. However at least as I understand it, Majordomo won't
work in this scenario.

Let's take the FIX protocol. Since FIX is asynchronous, I believe I can't
use the Majordomo. You may send an order to a FIX engine, and that one
order may send back many different messages. For example, that one order
may see the following responses: Insert, Pending, New, Partial Fill, then
finally Fill.

In the Majordomo pattern, a section of code in the worker looks like this

mdwrk session ("tcp://localhost:5555", sourceStr.c_str(), verbose);

zmsg *reply = 0;
while (1) {
    zmsg *request = session.recv (reply);
    if (request == 0) {
        break;              //  Worker was interrupted
    }

    //reply = request;        //  Echo is complex... :-)
    reply = new zmsg(sourceStr.c_str());
}


This assumes one request, and one reply. Since my application is not
strictly a single request, that gets a single response as in the code
above, do I have to drop Majordomo ?


Thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120726/2ba529a6/attachment.htm 

From ph at imatix.com  Fri Jul 27 07:58:24 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 27 Jul 2012 08:58:24 +0300
Subject: [zeromq-dev] [PATCH] tests: getting rid of warnings when build
 with NDEBUG.
In-Reply-To: <1343049340.16350.16.camel@vivanov>
References: <1343049340.16350.16.camel@vivanov>
Message-ID: <CADL5_sjztGN1CpiR8rCNZFc_9qPQqwppc++8uzgppQ2YXA2s_A@mail.gmail.com>

Hi Vitally,

Can you send any patch to 2.2 as a pull request, with a corresponding
issue in the issue tracker?

Thanks a lot
Pieter

On Mon, Jul 23, 2012 at 4:15 PM, Vitaliy Ivanov <vitaliy at toroki.com> wrote:
> Hi,
>
> ZMQ tree 2-x.
> I didn't use ZMQ_ASSERT as tests are not using this macro.
>
> Vitaliy
>
> ----------------------------------------------------------------------
>
> >From 12e7514e9c076f35cd525fee5f2c68a1228e53f2 Mon Sep 17 00:00:00 2001
> From: Vitaliy Ivanov <vitaliy at toroki.com>
> Date: Mon, 23 Jul 2012 16:06:13 +0300
> Subject: [PATCH] tests: getting rid of warnings when build with NDEBUG.
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
>
> Correcting the following warnings when build in non asserts mode:
> a.
> testutil.hpp: In function ?void zmqtestutil::basic_tests(const char*, int, int)?:
> testutil.hpp:114: error: unused variable ?rc?
> b.
> test_hwm.cpp: In function ?int main(int, char**)?:
> test_hwm.cpp:47: error: unused variable ?sent?
> test_hwm.cpp:60: error: unused variable ?received?
> test_hwm.cpp:64: error: unused variable ?sent?
>
> Signed-off-by: Vitaliy Ivanov <vitalivanov at gmail.com>
> ---
>  tests/test_hwm.cpp   |   17 +++++++++--------
>  tests/test_timeo.cpp |    4 ++--
>  tests/testutil.hpp   |   17 +++++++++--------
>  3 files changed, 20 insertions(+), 18 deletions(-)
>
> diff --git a/tests/test_hwm.cpp b/tests/test_hwm.cpp
> index d73c25f..e1fcec2 100644
> --- a/tests/test_hwm.cpp
> +++ b/tests/test_hwm.cpp
> @@ -28,6 +28,7 @@ int main (int argc, char *argv [])
>  {
>      uint64_t hwm = 5;
>      int linger = 0;
> +    bool rc;
>
>      zmq::context_t context (1);
>      zmq::socket_t s1 (context, ZMQ_PULL);
> @@ -44,25 +45,25 @@ int main (int argc, char *argv [])
>          zmq::message_t msg (sizeof ("test") - 1);
>          memcpy (msg.data (), "test", sizeof ("test") - 1);
>
> -        bool sent = s2.send (msg, ZMQ_NOBLOCK);
> +        rc = s2.send (msg, ZMQ_NOBLOCK);
>
>          // Anything below HWM should be sent
>          if (i < 5) {
> -            assert (sent);
> +            assert (rc);
>          } else {
> -            assert (!sent && errno == EAGAIN);
> +            assert (!rc && errno == EAGAIN);
>          }
>      }
>
>      // There should be now 5 messages pending, consume one
>      zmq::message_t msg;
>
> -    bool received = s1.recv (&msg, 0);
> -    assert (received);
> +    rc = s1.recv (&msg, 0);
> +    assert (rc);
>
>      // Now it should be possible to send one more
> -    bool sent = s2.send (msg, 0);
> -    assert (sent);
> +    rc = s2.send (msg, 0);
> +    assert (rc);
>
> -       return 0;
> +    return 0;
>  }
> diff --git a/tests/test_timeo.cpp b/tests/test_timeo.cpp
> index b9cd512..251972e 100644
> --- a/tests/test_timeo.cpp
> +++ b/tests/test_timeo.cpp
> @@ -107,7 +107,7 @@ int main (int argc, char *argv [])
>      memcpy (zmq_msg_data (&msg), "12345678ABCDEFGH12345678abcdefgh", 32);
>      rc = zmq_send (sc, &msg, 0);
>      assert (rc == 0);
> -
> +
>      rc = zmq_recv (sb, &msg, 0);
>      assert (rc == 0);
>      assert (32 == zmq_msg_size(&msg));
> @@ -120,6 +120,6 @@ int main (int argc, char *argv [])
>      rc = zmq_term (ctx);
>      assert (rc == 0);
>
> -    return 0 ;
> +    return 0;
>  }
>
> diff --git a/tests/testutil.hpp b/tests/testutil.hpp
> index 7f182c1..550ecf0 100644
> --- a/tests/testutil.hpp
> +++ b/tests/testutil.hpp
> @@ -30,8 +30,7 @@
>
>  namespace zmqtestutil
>  {
> -
> -    using namespace std ;
> +    using namespace std;
>
>      typedef std::pair <zmq::socket_t*, zmq::socket_t*> socket_pair;
>
> @@ -68,7 +67,7 @@ namespace zmqtestutil
>          s2.send (pong, 0);
>
>          //  Return received data as std::string.
> -        return ret ;
> +        return ret;
>      }
>
>      /*  Run basic tests for the given transport.
> @@ -79,6 +78,8 @@ namespace zmqtestutil
>      */
>      void basic_tests (const char *transport_, int t1_, int t2_)
>      {
> +        int rc = 0;
> +        string returned;
>          zmq::context_t context (1);
>
>          zmq::pollitem_t items [2];
> @@ -88,7 +89,7 @@ namespace zmqtestutil
>          const string expect ("XXX");
>
>          {
> -            const string returned = zmqtestutil::ping_pong (p, expect);
> +            returned = zmqtestutil::ping_pong (p, expect);
>              assert (expect == returned);
>
>              //  Adjust socket state so that poll shows only 1 pending message.
> @@ -97,7 +98,7 @@ namespace zmqtestutil
>          }
>
>          {
> -            //  Now poll is used to singal that a message is ready to read.
> +            //  Now poll is used to signal that a message is ready to read.
>              zmq::message_t m1 (expect.size ());
>              memcpy (m1.data (), expect.c_str (), expect.size ());
>              items [0].socket = *p.first;
> @@ -111,14 +112,14 @@ namespace zmqtestutil
>
>              p.first->send (m1, 0);
>
> -            int rc = zmq::poll (&items [0], 2, -1);
> +            rc = zmq::poll (&items [0], 2, -1);
>              assert (rc == 1);
>              assert ((items [1].revents & ZMQ_POLLIN) != 0);
>
>              zmq::message_t m2;
>              p.second->recv (&m2, 0);
> -            const string ret ((char*) m2.data (), m2.size ());
> -            assert (expect == ret);
> +            returned = string((char*) m2.data (), m2.size ());
> +            assert (expect == returned);
>          }
>
>          //  Delete sockets.
> --
> 1.7.0.4
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From ph at imatix.com  Fri Jul 27 08:33:08 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 27 Jul 2012 09:33:08 +0300
Subject: [zeromq-dev] Question about constness
In-Reply-To: <CAESgSZqOZGD+oNrzSuNF_JjH8sbqW0aZ1--VVu67=r=Y=scG1g@mail.gmail.com>
References: <500E28F6.1040608@gmail.com>
	<CAESgSZqOZGD+oNrzSuNF_JjH8sbqW0aZ1--VVu67=r=Y=scG1g@mail.gmail.com>
Message-ID: <CADL5_sivSuCgWDJ223ue0uNsRsgV67=KZe6kY+o5fmv-wxkTaA@mail.gmail.com>

On Tue, Jul 24, 2012 at 4:23 PM, Sebastian Lauwers
<sebastian.lauwers at gmail.com> wrote:

> Checking the zeromq 3.2 RC [1], I see this still hasn't been
> applied---I don't know what the project's policy would be regarding
> "breaking" the 3.2 API after the first RC has been released.
> Constness is used a handful of times in zmq.h.

Breaking the C API is a bad idea, no matter what the version numbering.

I'd suggest leaving the C API as is, and improving the C++ API (in
https://github.com/zeromq/cppzmq).

-Pieter

From sebastian.lauwers at gmail.com  Fri Jul 27 08:56:40 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Fri, 27 Jul 2012 08:56:40 +0200
Subject: [zeromq-dev] Question about constness
In-Reply-To: <CADL5_sivSuCgWDJ223ue0uNsRsgV67=KZe6kY+o5fmv-wxkTaA@mail.gmail.com>
References: <500E28F6.1040608@gmail.com>
	<CAESgSZqOZGD+oNrzSuNF_JjH8sbqW0aZ1--VVu67=r=Y=scG1g@mail.gmail.com>
	<CADL5_sivSuCgWDJ223ue0uNsRsgV67=KZe6kY+o5fmv-wxkTaA@mail.gmail.com>
Message-ID: <50123BA8.3080806@gmail.com>

On 27/07/12 08:33, Pieter Hintjens wrote:
> I'd suggest leaving the C API as is, and improving the C++ API (in
> https://github.com/zeromq/cppzmq).

Very well, I'll try to provide a patch tomorrow.

-S.

From ph at imatix.com  Fri Jul 27 09:03:26 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 27 Jul 2012 10:03:26 +0300
Subject: [zeromq-dev] [PATCH] [cppzmq] Add move constructor to message_t
In-Reply-To: <OF21D5BAA1.FE45B344-ON80257A47.00446023-80257A47.0044CCF8@waters.com>
References: <OF21D5BAA1.FE45B344-ON80257A47.00446023-80257A47.0044CCF8@waters.com>
Message-ID: <CADL5_siQi69HEojq8etz5QrbPtgzG_idDKS9Rba5Nvrd=dHX2Q@mail.gmail.com>

Hi Richard,

The process for contributions is described here:
http://www.zeromq.org/docs:contributing

Basically you want to make github pull requests against the cppzmq project.

-Pieter

On Thu, Jul 26, 2012 at 3:31 PM,  <Richard_Newton at waters.com> wrote:
> Hi,
>
> In the C++ binding both context_t and socket_t have move constructors, I
> would like to extend that to message_t as well.
>
>
> (See attached file: Add-move-constructor-to-message_t.patch)
>
> This was again the master branch on github, also sent corresponding pull
> request.
>
> Regards,
>
> Richard.
>
> ===========================================================
> The information in this email is confidential, and is intended solely for
> the addressee(s).
> Access to this email by anyone else is unauthorized and therefore
> prohibited.  If you are
> not the intended recipient you are notified that disclosing, copying,
> distributing or taking
> any action in reliance on the contents of this information is strictly
> prohibited and may be unlawful.
> ===========================================================
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From ph at imatix.com  Fri Jul 27 09:07:33 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Fri, 27 Jul 2012 10:07:33 +0300
Subject: [zeromq-dev] If not Majordomo for asynchronous workers,
	then what?
In-Reply-To: <CAEiCcQBXNrGyWqj17tgTDzJw4gVqmHp7Vj-QoJYj4zNYnaMgow@mail.gmail.com>
References: <CAEiCcQBXNrGyWqj17tgTDzJw4gVqmHp7Vj-QoJYj4zNYnaMgow@mail.gmail.com>
Message-ID: <CADL5_sg5qBPbqZaQVBS_Z_WD_fN2WwvtJbxOxaNGhSaTcGJ5kg@mail.gmail.com>

Hi Ivan,

Majordomo is based on a 1-to-1 request-reply model for workers, yes.
You should be able to quite easily modify MDP to allow multiple
replies, perhaps by adding a "there's more coming" indicator to
replies that the broker uses to route several replies back to the same
client.

Modifying MDP and the implementations will be simpler than writing
your own from scratch.

You can also use other architectures; MDP is specifically aimed at
allowing multiple redundant workers per service type. If you have a
single well-known gateway per exchange you can just connect each
client to each exchange gateway directly and work fully
asynchronously.

-Pieter

On Thu, Jul 26, 2012 at 10:29 PM, Ivan Figueredo <ivan at blackpx.com> wrote:
> I am writing a distributed trading manager. There are many different
> liquidity providers, and the client would like to route to a given exchange.
> With this in mind, I thought that using the Majordomo pattern with a broker
> in between clients and workers asking a particular [named] worker to do its
> work. However at least as I understand it, Majordomo won't work in this
> scenario.
>
> Let's take the FIX protocol. Since FIX is asynchronous, I believe I can't
> use the Majordomo. You may send an order to a FIX engine, and that one order
> may send back many different messages. For example, that one order may see
> the following responses: Insert, Pending, New, Partial Fill, then finally
> Fill.
>
> In the Majordomo pattern, a section of code in the worker looks like this
>
> mdwrk session ("tcp://localhost:5555", sourceStr.c_str(), verbose);
>
> zmsg *reply = 0;
> while (1) {
>     zmsg *request = session.recv (reply);
>     if (request == 0) {
>         break;              //  Worker was interrupted
>     }
>
>     //reply = request;        //  Echo is complex... :-)
>     reply = new zmsg(sourceStr.c_str());
> }
>
>
> This assumes one request, and one reply. Since my application is not
> strictly a single request, that gets a single response as in the code above,
> do I have to drop Majordomo ?
>
>
> Thanks!
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>

From jonas.adler at epiq.se  Fri Jul 27 09:08:39 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Fri, 27 Jul 2012 09:08:39 +0200 (CEST)
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
Message-ID: <46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>

If i manually type in "arm-linux-androideabi-" and press tab twice i get a
list of all the compilers (gcc, g++ etc), so it must be in my path.

Also the command "arm-linux-androideabi-gcc" is executable.

> Seems there is an issue in your PATH. From the line 83 (and the others) it
> looks like the toolchain is not found:
>
> configure:3246: checking for arm-linux-androideabi-gcc
> configure:3276: result: no
> configure:3286: checking for gcc
> configure:3302: found /usr/bin/gcc
> configure:3313: result: gcc
>
> ... and so on. In the end the compiler complains about clock_gettime() and
> pthread_create not found, but that;s an aftereffect of not using the right
> compiler.
>
> Could you check your $PATH ?
>
> On Wed, Jul 25, 2012 at 4:20 PM, <jonas.adler at epiq.se> wrote:
>
>> Oh i forgot to add the link to the log, I'm sorry :p
>>
>> http://pastebin.com/pWwepPdr
>>
>> > Cqn you please send me your config.log ?
>> >
>> > BR,
>> >
>> > On Wed, Jul 25, 2012 at 3:17 PM, <jonas.adler at epiq.se> wrote:
>> >
>> >> I still get the same problem with zmq_init not in -lzmq.
>> >>
>> >> some enviroment variables that might be useful:
>> >>
>> >>
>> >>
>> PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games
>> >>
>> >> OUTPUT_DIR=/tmp/zeromq-android
>> >>
>> >> packages I installed (running ubuntu x86_64 clean install before i
>> >> tried)
>> >> was:
>> >>  * libtool, autogen, automake
>> >>  * gcc, gpp, gcc-multilib
>> >>  * libc6-dev
>> >>  * git
>> >>
>> >> something I'm missing?
>> >>
>> >> The tarball for zmq I downloaded was
>> >> http://download.zeromq.org/zeromq-2.2.0.tar.gz
>> >> (with wget into /tmp/. Since the wiki doesn't state where to get it i
>> >> just
>> >> took it from the download page)
>> >>
>> >> Also here is the config.log file the JZMQ ./config generated (with
>> that
>> >> LIBS"-luuid" added to the flags, as you can see on line 7)
>> >>
>> >> Thanks in advance.
>> >>
>> >> > OK, I ran the exact same sequence. Seems you are right, I have
>> >> forgotten
>> >> a
>> >> > little word on the last configure line.
>> >> > When it fails, the config.log says:
>> >> >
>> >> > 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function
>> >> `uuid_t':
>> >> > 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
>> >> > `uuid_generate'
>> >> > 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
>> >> > `uuid_unparse'
>> >> > 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
>> >> > `uuid_generate'
>> >> > 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
>> >> > `uuid_unparse'
>> >> >
>> >> > Well if you recompile jzmq with the line:
>> >> >
>> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
>> >> > LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
>> >> >
>> >> > and everything seems fine. Can you confirm it works ?
>> >> >
>> >> > Best regards,
>> >> >
>> >> > On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
>> >> >
>> >> >> Yes, I configured zmq before I tried to configure jzmq.
>> >> >> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured
>> zmq.
>> >> >>
>> >> >> Here's a link to the script I made (from the instructions on
>> >> >> build:android) after redoing the instructions approximate 15
>> times.
>> >> >>
>> >> >> Worth noting is that I installed the NDK and added the
>> >> >> /opt/android-toolchain/bin to my PATH by hand, the steps i
>> followed
>> >> are
>> >> >> at
>> >> >> the top of this pastebin post.
>> >> >>
>> >> >> http://pastebin.com/iXQWCxNV
>> >> >>
>> >> >> Everything seems to go smooth untill the jzmq config. That is
>> where i
>> >> >> get
>> >> >> the "configure: error: cannot link with -lzmq" error.
>> >> >>
>> >> >> Everything I did tried I did with sudo or root access.
>> >> >>
>> >> >> > Did you try to configure zmq or jzmq?The configure command you
>> >> >> described
>> >> >> > is command line for jzmq. You need first to have set the env
>> >> >> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated
>> >> zmq.
>> >> >> > Then you should be able to launch configure for jzmq.
>> >> >> > Regards.
>> >> >> >
>> >> >> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
>> >> >> <jonas.adler at epiq.se>
>> >> >> a
>> >> >> > ?crit :
>> >> >> >
>> >> >> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
>> >> >> > Objet: [zeromq-dev] Massive problems running ZMQ on android
>> >> >> > ?: zeromq-dev at lists.zeromq.org
>> >> >> > Date: Mardi 24 juillet 2012, 16h03
>> >> >> >
>> >> >> > Hello,
>> >> >> >
>> >> >> > I'm new to ZMQ, and trying to get it to run on android. I am
>> >> however,
>> >> >> > facing issue, after issue, after issue.
>> >> >> >
>> >> >> > I try to follow the official guide,
>> >> >> (http://www.zeromq.org/build:android
>> >> >> ),
>> >> >> > and have gotten to the following call on my ubuntu machine:
>> >> >> >
>> >> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> >> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
>> >> >> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
>> >> >> >
>> >> >> > I get this error:
>> >> >> >
>> >> >> > "configure: error: cannot link with -lzmq"
>> >> >> >
>> >> >> > I have tried google for a few days now, but i cannot find any
>> >> solution
>> >> >> to
>> >> >> > this that works for me.
>> >> >> >
>> >> >> > Some comments on other solutions i have found, that did not work
>> >> for
>> >> >> me:
>> >> >> >
>> >> >> > I have installed pretty much every compiler known to man, but if
>> >> you
>> >> >> have
>> >> >> > any specific suggestions, please tell me.
>> >> >> > I have made no alterations to the source at all, simply
>> following
>> >> the
>> >> >> > instructions to the letter.
>> >> >> >
>> >> >> > Has someone else had this error? Is there something that is
>> >> >> "understood"
>> >> >> > in the guide, that i could be missing?
>> >> >> >
>> >> >> > Any help would be greatly appreciated!
>> >> >> >
>> >> >> > //Jonas Adler
>> >> >> >
>> >> >> > _______________________________________________
>> >> >> > zeromq-dev mailing list
>> >> >> > zeromq-dev at lists.zeromq.org
>> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >> > _______________________________________________
>> >> >> > zeromq-dev mailing list
>> >> >> > zeromq-dev at lists.zeromq.org
>> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >> >
>> >> >>
>> >> >>
>> >> >
>> >> >
>> >> > --
>> >> > Victor
>> >> > _______________________________________________
>> >> > zeromq-dev mailing list
>> >> > zeromq-dev at lists.zeromq.org
>> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >
>> >>
>> >>
>> >
>> >
>> > --
>> > Victor
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>>
>>
>>
>
>
> --
> Victor
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



From victor at iso3103.net  Fri Jul 27 09:30:21 2012
From: victor at iso3103.net (Victor Perron)
Date: Fri, 27 Jul 2012 09:30:21 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
Message-ID: <CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>

I don't know what happens then with your configuration. Again, from your
config.log:

   1.
   2. PATH: /usr/local/sbin
   3. PATH: /usr/local/bin
   4. PATH: /usr/sbin
   5. PATH: /usr/bin
   6. PATH: /sbin
   7. PATH: /bin
   8.
   9.
   10. ## ----------- ##
   11. ## Core tests. ##
   12. ## ----------- ##
   13.
   14. configure:2454: checking for a BSD-compatible install
   15. configure:2522: result: /usr/bin/install -c
   16. configure:2533: checking whether build environment is sane
   17. configure:2583: result: yes
   18. *configure:2632: checking for arm-linux-androideabi-strip*
   19. *configure:2662: result: no*
   20. configure:2672: checking for strip
   21. configure:2688: found /usr/bin/strip
   22. configure:2699: result: strip
   23. configure:2724: checking for a thread-safe mkdir -p
   24. configure:2763: result: /bin/mkdir -p
   25. configure:2776: checking for gawk
   26. configure:2806: result: no
   27. configure:2776: checking for mawk
   28. configure:2792: found /usr/bin/mawk
   29. configure:2803: result: mawk
   30. configure:2814: checking whether make sets $(MAKE)
   31. configure:2836: result: yes
   32. configure:2911: checking how to create a ustar tar archive
   33. configure:2924: tar --version
   34. tar (GNU tar) 1.26
   35. Copyright (C) 2011 Free Software Foundation, Inc.
   36. License GPLv3+: GNU GPL version 3 or later <
   http://gnu.org/licenses/gpl.html>.
   37. This is free software: you are free to change and redistribute it.
   38. There is NO WARRANTY, to the extent permitted by law.
   39.
   40. Written by John Gilmore and Jay Fenlason.
   41. configure:2927: $? = 0
   42. configure:2967: tardir=conftest.dir && eval tar --format=ustar -chf
   - "$tardir" >conftest.tar
   43. configure:2970: $? = 0
   44. configure:2974: tar -xf - <conftest.tar
   45. configure:2977: $? = 0
   46. configure:2990: result: gnutar
   47. configure:3039: checking build system type
   48. configure:3053: result: x86_64-unknown-linux-gnu
   49. configure:3073: checking host system type
   50. configure:3086: result: arm-unknown-linux-androideabi
   51. configure:3127: checking how to print strings
   52. configure:3154: result: printf
   53. configure:3187: checking for style of include used by make
   54. configure:3215: result: GNU
   55. configure:3246: checking for arm-linux-androideabi-gcc
   56. configure:3276: result: no
   57. configure:3286: checking for gcc
   58. configure:3302: found /usr/bin/gcc
   59. configure:3313: result: gcc
   60. configure:3542: checking for C compiler version
   61. configure:3551: gcc --version >&5
   62. gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
   63. Copyright (C) 2011 Free Software Foundation, Inc.
   64. This is free software; see the source for copying conditions.  There
   is NO
   65. warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR
   PURPOSE.
   66.


My config.log with the exact same autogen/configure/... suite :

PATH: /opt/swig-2.0.7/bin
PATH: /home/victor/work/bin/arm-linux-androideabi-4.4.3/bin
PATH: /home/victor/work/bin/apache-maven-3.0.4/bin
PATH: /home/victor/work/bin/jdk1.6.0_30/bin
PATH: /home/victor/work/bin/android-ndk-r8
PATH: /home/victor/work/bin/android-sdk-linux/tools
PATH: /home/victor/work/bin/android-sdk-linux/platform-tools
PATH: /home/victor/work/bin/eclipse
PATH: /home/victor/bin
PATH: /home/victor/work/bin
PATH: /home/victor/bin
PATH: /usr/local/sbin
PATH: /usr/local/bin
PATH: /usr/sbin
PATH: /usr/bin
PATH: /sbin
PATH: /bin
PATH: /usr/games


## ----------- ##
## Core tests. ##
## ----------- ##

configure:2454: checking for a BSD-compatible install
configure:2522: result: /usr/bin/install -c
configure:2533: checking whether build environment is sane
configure:2583: result: yes
configure:2632: checking for arm-linux-androideabi-strip
configure:2648: found
/home/victor/work/bin/arm-linux-androideabi-4.4.3/bin/arm-linux-androideabi-strip
configure:2659: result: arm-linux-androideabi-strip
configure:2724: checking for a thread-safe mkdir -p
configure:2763: result: /bin/mkdir -p
configure:2776: checking for gawk
configure:2792: found /usr/bin/gawk
configure:2803: result: gawk
configure:2814: checking whether make sets $(MAKE)
configure:2836: result: yes
configure:2911: checking how to create a ustar tar archive
configure:2924: tar --version
tar (GNU tar) 1.26
Copyright (C) 2011 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html
>.
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

Written by John Gilmore and Jay Fenlason.
configure:2927: $? = 0
configure:2967: tardir=conftest.dir && eval tar --format=ustar -chf -
"$tardir" >conftest.tar
configure:2970: $? = 0
configure:2974: tar -xf - <conftest.tar
configure:2977: $? = 0
configure:2990: result: gnutar
configure:3039: checking build system type
configure:3053: result: x86_64-unknown-linux-gnu
configure:3073: checking host system type
configure:3086: result: arm-unknown-linux-androideabi
configure:3127: checking how to print strings
configure:3154: result: printf
configure:3187: checking for style of include used by make
configure:3215: result: GNU
configure:3246: checking for arm-linux-androideabi-gcc
configure:3262: found
/home/victor/work/bin/arm-linux-androideabi-4.4.3/bin/arm-linux-androideabi-gcc
configure:3273: result: arm-linux-androideabi-gcc
configure:3542: checking for C compiler version
configure:3551: arm-linux-androideabi-gcc --version >&5
arm-linux-androideabi-gcc (GCC) 4.4.3
Copyright (C) 2009 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.


That is quite a difference ! :)


On Fri, Jul 27, 2012 at 9:08 AM, <jonas.adler at epiq.se> wrote:

> If i manually type in "arm-linux-androideabi-" and press tab twice i get a
> list of all the compilers (gcc, g++ etc), so it must be in my path.
>
> Also the command "arm-linux-androideabi-gcc" is executable.
>
> > Seems there is an issue in your PATH. From the line 83 (and the others)
> it
> > looks like the toolchain is not found:
> >
> > configure:3246: checking for arm-linux-androideabi-gcc
> > configure:3276: result: no
> > configure:3286: checking for gcc
> > configure:3302: found /usr/bin/gcc
> > configure:3313: result: gcc
> >
> > ... and so on. In the end the compiler complains about clock_gettime()
> and
> > pthread_create not found, but that;s an aftereffect of not using the
> right
> > compiler.
> >
> > Could you check your $PATH ?
> >
> > On Wed, Jul 25, 2012 at 4:20 PM, <jonas.adler at epiq.se> wrote:
> >
> >> Oh i forgot to add the link to the log, I'm sorry :p
> >>
> >> http://pastebin.com/pWwepPdr
> >>
> >> > Cqn you please send me your config.log ?
> >> >
> >> > BR,
> >> >
> >> > On Wed, Jul 25, 2012 at 3:17 PM, <jonas.adler at epiq.se> wrote:
> >> >
> >> >> I still get the same problem with zmq_init not in -lzmq.
> >> >>
> >> >> some enviroment variables that might be useful:
> >> >>
> >> >>
> >> >>
> >>
> PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games
> >> >>
> >> >> OUTPUT_DIR=/tmp/zeromq-android
> >> >>
> >> >> packages I installed (running ubuntu x86_64 clean install before i
> >> >> tried)
> >> >> was:
> >> >>  * libtool, autogen, automake
> >> >>  * gcc, gpp, gcc-multilib
> >> >>  * libc6-dev
> >> >>  * git
> >> >>
> >> >> something I'm missing?
> >> >>
> >> >> The tarball for zmq I downloaded was
> >> >> http://download.zeromq.org/zeromq-2.2.0.tar.gz
> >> >> (with wget into /tmp/. Since the wiki doesn't state where to get it i
> >> >> just
> >> >> took it from the download page)
> >> >>
> >> >> Also here is the config.log file the JZMQ ./config generated (with
> >> that
> >> >> LIBS"-luuid" added to the flags, as you can see on line 7)
> >> >>
> >> >> Thanks in advance.
> >> >>
> >> >> > OK, I ran the exact same sequence. Seems you are right, I have
> >> >> forgotten
> >> >> a
> >> >> > little word on the last configure line.
> >> >> > When it fails, the config.log says:
> >> >> >
> >> >> > 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function
> >> >> `uuid_t':
> >> >> > 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> >> >> > `uuid_generate'
> >> >> > 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> >> >> > `uuid_unparse'
> >> >> > 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
> >> >> > `uuid_generate'
> >> >> > 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
> >> >> > `uuid_unparse'
> >> >> >
> >> >> > Well if you recompile jzmq with the line:
> >> >> >
> >> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> >> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> >> >> > LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
> >> >> >
> >> >> > and everything seems fine. Can you confirm it works ?
> >> >> >
> >> >> > Best regards,
> >> >> >
> >> >> > On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
> >> >> >
> >> >> >> Yes, I configured zmq before I tried to configure jzmq.
> >> >> >> My OUTPUT_DIR was set to /tmp/zeromq-android before I configured
> >> zmq.
> >> >> >>
> >> >> >> Here's a link to the script I made (from the instructions on
> >> >> >> build:android) after redoing the instructions approximate 15
> >> times.
> >> >> >>
> >> >> >> Worth noting is that I installed the NDK and added the
> >> >> >> /opt/android-toolchain/bin to my PATH by hand, the steps i
> >> followed
> >> >> are
> >> >> >> at
> >> >> >> the top of this pastebin post.
> >> >> >>
> >> >> >> http://pastebin.com/iXQWCxNV
> >> >> >>
> >> >> >> Everything seems to go smooth untill the jzmq config. That is
> >> where i
> >> >> >> get
> >> >> >> the "configure: error: cannot link with -lzmq" error.
> >> >> >>
> >> >> >> Everything I did tried I did with sudo or root access.
> >> >> >>
> >> >> >> > Did you try to configure zmq or jzmq?The configure command you
> >> >> >> described
> >> >> >> > is command line for jzmq. You need first to have set the env
> >> >> >> > variable OUTPUT_DIR=/tmp/zeromq-android and then have generated
> >> >> zmq.
> >> >> >> > Then you should be able to launch configure for jzmq.
> >> >> >> > Regards.
> >> >> >> >
> >> >> >> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
> >> >> >> <jonas.adler at epiq.se>
> >> >> >> a
> >> >> >> > ?crit :
> >> >> >> >
> >> >> >> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
> >> >> >> > Objet: [zeromq-dev] Massive problems running ZMQ on android
> >> >> >> > ?: zeromq-dev at lists.zeromq.org
> >> >> >> > Date: Mardi 24 juillet 2012, 16h03
> >> >> >> >
> >> >> >> > Hello,
> >> >> >> >
> >> >> >> > I'm new to ZMQ, and trying to get it to run on android. I am
> >> >> however,
> >> >> >> > facing issue, after issue, after issue.
> >> >> >> >
> >> >> >> > I try to follow the official guide,
> >> >> >> (http://www.zeromq.org/build:android
> >> >> >> ),
> >> >> >> > and have gotten to the following call on my ubuntu machine:
> >> >> >> >
> >> >> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
> >> >> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
> >> >> >> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
> >> >> >> >
> >> >> >> > I get this error:
> >> >> >> >
> >> >> >> > "configure: error: cannot link with -lzmq"
> >> >> >> >
> >> >> >> > I have tried google for a few days now, but i cannot find any
> >> >> solution
> >> >> >> to
> >> >> >> > this that works for me.
> >> >> >> >
> >> >> >> > Some comments on other solutions i have found, that did not work
> >> >> for
> >> >> >> me:
> >> >> >> >
> >> >> >> > I have installed pretty much every compiler known to man, but if
> >> >> you
> >> >> >> have
> >> >> >> > any specific suggestions, please tell me.
> >> >> >> > I have made no alterations to the source at all, simply
> >> following
> >> >> the
> >> >> >> > instructions to the letter.
> >> >> >> >
> >> >> >> > Has someone else had this error? Is there something that is
> >> >> >> "understood"
> >> >> >> > in the guide, that i could be missing?
> >> >> >> >
> >> >> >> > Any help would be greatly appreciated!
> >> >> >> >
> >> >> >> > //Jonas Adler
> >> >> >> >
> >> >> >> > _______________________________________________
> >> >> >> > zeromq-dev mailing list
> >> >> >> > zeromq-dev at lists.zeromq.org
> >> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >> >> > _______________________________________________
> >> >> >> > zeromq-dev mailing list
> >> >> >> > zeromq-dev at lists.zeromq.org
> >> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >> >> >
> >> >> >>
> >> >> >>
> >> >> >
> >> >> >
> >> >> > --
> >> >> > Victor
> >> >> > _______________________________________________
> >> >> > zeromq-dev mailing list
> >> >> > zeromq-dev at lists.zeromq.org
> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >> >
> >> >>
> >> >>
> >> >
> >> >
> >> > --
> >> > Victor
> >> > _______________________________________________
> >> > zeromq-dev mailing list
> >> > zeromq-dev at lists.zeromq.org
> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >
> >>
> >>
> >>
> >
> >
> > --
> > Victor
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/293132e7/attachment.htm 

From jonas.adler at epiq.se  Fri Jul 27 09:43:54 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Fri, 27 Jul 2012 09:43:54 +0200 (CEST)
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
	<CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
Message-ID: <c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>

I think i've found something. If I use sudo <configure> i get that
arm-linux-androideabi-gcc isnt found, but when i dont use sudo and it is
found, i get "C compiler cannot create executables"...

arm-linux-androideabi-gcc -> works
sudo arm-linux-androideabi-gcc -> command not found

> I don't know what happens then with your configuration. Again, from your
> config.log:
>
>    1.
>    2. PATH: /usr/local/sbin
>    3. PATH: /usr/local/bin
>    4. PATH: /usr/sbin
>    5. PATH: /usr/bin
>    6. PATH: /sbin
>    7. PATH: /bin
>    8.
>    9.
>    10. ## ----------- ##
>    11. ## Core tests. ##
>    12. ## ----------- ##
>    13.
>    14. configure:2454: checking for a BSD-compatible install
>    15. configure:2522: result: /usr/bin/install -c
>    16. configure:2533: checking whether build environment is sane
>    17. configure:2583: result: yes
>    18. *configure:2632: checking for arm-linux-androideabi-strip*
>    19. *configure:2662: result: no*
>    20. configure:2672: checking for strip
>    21. configure:2688: found /usr/bin/strip
>    22. configure:2699: result: strip
>    23. configure:2724: checking for a thread-safe mkdir -p
>    24. configure:2763: result: /bin/mkdir -p
>    25. configure:2776: checking for gawk
>    26. configure:2806: result: no
>    27. configure:2776: checking for mawk
>    28. configure:2792: found /usr/bin/mawk
>    29. configure:2803: result: mawk
>    30. configure:2814: checking whether make sets $(MAKE)
>    31. configure:2836: result: yes
>    32. configure:2911: checking how to create a ustar tar archive
>    33. configure:2924: tar --version
>    34. tar (GNU tar) 1.26
>    35. Copyright (C) 2011 Free Software Foundation, Inc.
>    36. License GPLv3+: GNU GPL version 3 or later <
>    http://gnu.org/licenses/gpl.html>.
>    37. This is free software: you are free to change and redistribute it.
>    38. There is NO WARRANTY, to the extent permitted by law.
>    39.
>    40. Written by John Gilmore and Jay Fenlason.
>    41. configure:2927: $? = 0
>    42. configure:2967: tardir=conftest.dir && eval tar --format=ustar -chf
>    - "$tardir" >conftest.tar
>    43. configure:2970: $? = 0
>    44. configure:2974: tar -xf - <conftest.tar
>    45. configure:2977: $? = 0
>    46. configure:2990: result: gnutar
>    47. configure:3039: checking build system type
>    48. configure:3053: result: x86_64-unknown-linux-gnu
>    49. configure:3073: checking host system type
>    50. configure:3086: result: arm-unknown-linux-androideabi
>    51. configure:3127: checking how to print strings
>    52. configure:3154: result: printf
>    53. configure:3187: checking for style of include used by make
>    54. configure:3215: result: GNU
>    55. configure:3246: checking for arm-linux-androideabi-gcc
>    56. configure:3276: result: no
>    57. configure:3286: checking for gcc
>    58. configure:3302: found /usr/bin/gcc
>    59. configure:3313: result: gcc
>    60. configure:3542: checking for C compiler version
>    61. configure:3551: gcc --version >&5
>    62. gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
>    63. Copyright (C) 2011 Free Software Foundation, Inc.
>    64. This is free software; see the source for copying conditions.
> There
>    is NO
>    65. warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR
>    PURPOSE.
>    66.
>
>
> My config.log with the exact same autogen/configure/... suite :
>
> PATH: /opt/swig-2.0.7/bin
> PATH: /home/victor/work/bin/arm-linux-androideabi-4.4.3/bin
> PATH: /home/victor/work/bin/apache-maven-3.0.4/bin
> PATH: /home/victor/work/bin/jdk1.6.0_30/bin
> PATH: /home/victor/work/bin/android-ndk-r8
> PATH: /home/victor/work/bin/android-sdk-linux/tools
> PATH: /home/victor/work/bin/android-sdk-linux/platform-tools
> PATH: /home/victor/work/bin/eclipse
> PATH: /home/victor/bin
> PATH: /home/victor/work/bin
> PATH: /home/victor/bin
> PATH: /usr/local/sbin
> PATH: /usr/local/bin
> PATH: /usr/sbin
> PATH: /usr/bin
> PATH: /sbin
> PATH: /bin
> PATH: /usr/games
>
>
> ## ----------- ##
> ## Core tests. ##
> ## ----------- ##
>
> configure:2454: checking for a BSD-compatible install
> configure:2522: result: /usr/bin/install -c
> configure:2533: checking whether build environment is sane
> configure:2583: result: yes
> configure:2632: checking for arm-linux-androideabi-strip
> configure:2648: found
> /home/victor/work/bin/arm-linux-androideabi-4.4.3/bin/arm-linux-androideabi-strip
> configure:2659: result: arm-linux-androideabi-strip
> configure:2724: checking for a thread-safe mkdir -p
> configure:2763: result: /bin/mkdir -p
> configure:2776: checking for gawk
> configure:2792: found /usr/bin/gawk
> configure:2803: result: gawk
> configure:2814: checking whether make sets $(MAKE)
> configure:2836: result: yes
> configure:2911: checking how to create a ustar tar archive
> configure:2924: tar --version
> tar (GNU tar) 1.26
> Copyright (C) 2011 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later
> <http://gnu.org/licenses/gpl.html
>>.
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.
>
> Written by John Gilmore and Jay Fenlason.
> configure:2927: $? = 0
> configure:2967: tardir=conftest.dir && eval tar --format=ustar -chf -
> "$tardir" >conftest.tar
> configure:2970: $? = 0
> configure:2974: tar -xf - <conftest.tar
> configure:2977: $? = 0
> configure:2990: result: gnutar
> configure:3039: checking build system type
> configure:3053: result: x86_64-unknown-linux-gnu
> configure:3073: checking host system type
> configure:3086: result: arm-unknown-linux-androideabi
> configure:3127: checking how to print strings
> configure:3154: result: printf
> configure:3187: checking for style of include used by make
> configure:3215: result: GNU
> configure:3246: checking for arm-linux-androideabi-gcc
> configure:3262: found
> /home/victor/work/bin/arm-linux-androideabi-4.4.3/bin/arm-linux-androideabi-gcc
> configure:3273: result: arm-linux-androideabi-gcc
> configure:3542: checking for C compiler version
> configure:3551: arm-linux-androideabi-gcc --version >&5
> arm-linux-androideabi-gcc (GCC) 4.4.3
> Copyright (C) 2009 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.  There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR
> PURPOSE.
>
>
> That is quite a difference ! :)
>
>
> On Fri, Jul 27, 2012 at 9:08 AM, <jonas.adler at epiq.se> wrote:
>
>> If i manually type in "arm-linux-androideabi-" and press tab twice i get
>> a
>> list of all the compilers (gcc, g++ etc), so it must be in my path.
>>
>> Also the command "arm-linux-androideabi-gcc" is executable.
>>
>> > Seems there is an issue in your PATH. From the line 83 (and the
>> others)
>> it
>> > looks like the toolchain is not found:
>> >
>> > configure:3246: checking for arm-linux-androideabi-gcc
>> > configure:3276: result: no
>> > configure:3286: checking for gcc
>> > configure:3302: found /usr/bin/gcc
>> > configure:3313: result: gcc
>> >
>> > ... and so on. In the end the compiler complains about clock_gettime()
>> and
>> > pthread_create not found, but that;s an aftereffect of not using the
>> right
>> > compiler.
>> >
>> > Could you check your $PATH ?
>> >
>> > On Wed, Jul 25, 2012 at 4:20 PM, <jonas.adler at epiq.se> wrote:
>> >
>> >> Oh i forgot to add the link to the log, I'm sorry :p
>> >>
>> >> http://pastebin.com/pWwepPdr
>> >>
>> >> > Cqn you please send me your config.log ?
>> >> >
>> >> > BR,
>> >> >
>> >> > On Wed, Jul 25, 2012 at 3:17 PM, <jonas.adler at epiq.se> wrote:
>> >> >
>> >> >> I still get the same problem with zmq_init not in -lzmq.
>> >> >>
>> >> >> some enviroment variables that might be useful:
>> >> >>
>> >> >>
>> >> >>
>> >>
>> PATH=/opt/android-toolchain/bin:/use/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:usr/games
>> >> >>
>> >> >> OUTPUT_DIR=/tmp/zeromq-android
>> >> >>
>> >> >> packages I installed (running ubuntu x86_64 clean install before i
>> >> >> tried)
>> >> >> was:
>> >> >>  * libtool, autogen, automake
>> >> >>  * gcc, gpp, gcc-multilib
>> >> >>  * libc6-dev
>> >> >>  * git
>> >> >>
>> >> >> something I'm missing?
>> >> >>
>> >> >> The tarball for zmq I downloaded was
>> >> >> http://download.zeromq.org/zeromq-2.2.0.tar.gz
>> >> >> (with wget into /tmp/. Since the wiki doesn't state where to get
>> it i
>> >> >> just
>> >> >> took it from the download page)
>> >> >>
>> >> >> Also here is the config.log file the JZMQ ./config generated (with
>> >> that
>> >> >> LIBS"-luuid" added to the flags, as you can see on line 7)
>> >> >>
>> >> >> Thanks in advance.
>> >> >>
>> >> >> > OK, I ran the exact same sequence. Seems you are right, I have
>> >> >> forgotten
>> >> >> a
>> >> >> > little word on the last configure line.
>> >> >> > When it fails, the config.log says:
>> >> >> >
>> >> >> > 504 /tmp/myandroid/lib/libzmq.a(libzmq_la-uuid.o): In function
>> >> >> `uuid_t':
>> >> >> > 505 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
>> >> >> > `uuid_generate'
>> >> >> > 506 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
>> >> >> > `uuid_unparse'
>> >> >> > 507 /tmp/zeromq-2.2.0/src/uuid.cpp:88: undefined reference to
>> >> >> > `uuid_generate'
>> >> >> > 508 /tmp/zeromq-2.2.0/src/uuid.cpp:89: undefined reference to
>> >> >> > `uuid_unparse'
>> >> >> >
>> >> >> > Well if you recompile jzmq with the line:
>> >> >> >
>> >> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> >> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC -I$OUTPUT_DIR/include"
>> >> >> > LDFLAGS="-L$OUTPUT_DIR/lib"* LIBS="-luuid" *--disable-version
>> >> >> >
>> >> >> > and everything seems fine. Can you confirm it works ?
>> >> >> >
>> >> >> > Best regards,
>> >> >> >
>> >> >> > On Wed, Jul 25, 2012 at 1:30 PM, <jonas.adler at epiq.se> wrote:
>> >> >> >
>> >> >> >> Yes, I configured zmq before I tried to configure jzmq.
>> >> >> >> My OUTPUT_DIR was set to /tmp/zeromq-android before I
>> configured
>> >> zmq.
>> >> >> >>
>> >> >> >> Here's a link to the script I made (from the instructions on
>> >> >> >> build:android) after redoing the instructions approximate 15
>> >> times.
>> >> >> >>
>> >> >> >> Worth noting is that I installed the NDK and added the
>> >> >> >> /opt/android-toolchain/bin to my PATH by hand, the steps i
>> >> followed
>> >> >> are
>> >> >> >> at
>> >> >> >> the top of this pastebin post.
>> >> >> >>
>> >> >> >> http://pastebin.com/iXQWCxNV
>> >> >> >>
>> >> >> >> Everything seems to go smooth untill the jzmq config. That is
>> >> where i
>> >> >> >> get
>> >> >> >> the "configure: error: cannot link with -lzmq" error.
>> >> >> >>
>> >> >> >> Everything I did tried I did with sudo or root access.
>> >> >> >>
>> >> >> >> > Did you try to configure zmq or jzmq?The configure command
>> you
>> >> >> >> described
>> >> >> >> > is command line for jzmq. You need first to have set the env
>> >> >> >> > variable OUTPUT_DIR=/tmp/zeromq-android and then have
>> generated
>> >> >> zmq.
>> >> >> >> > Then you should be able to launch configure for jzmq.
>> >> >> >> > Regards.
>> >> >> >> >
>> >> >> >> > --- En date de : Mar 24.7.12, jonas.adler at epiq.se
>> >> >> >> <jonas.adler at epiq.se>
>> >> >> >> a
>> >> >> >> > ?crit :
>> >> >> >> >
>> >> >> >> > De: jonas.adler at epiq.se <jonas.adler at epiq.se>
>> >> >> >> > Objet: [zeromq-dev] Massive problems running ZMQ on android
>> >> >> >> > ?: zeromq-dev at lists.zeromq.org
>> >> >> >> > Date: Mardi 24 juillet 2012, 16h03
>> >> >> >> >
>> >> >> >> > Hello,
>> >> >> >> >
>> >> >> >> > I'm new to ZMQ, and trying to get it to run on android. I am
>> >> >> however,
>> >> >> >> > facing issue, after issue, after issue.
>> >> >> >> >
>> >> >> >> > I try to follow the official guide,
>> >> >> >> (http://www.zeromq.org/build:android
>> >> >> >> ),
>> >> >> >> > and have gotten to the following call on my ubuntu machine:
>> >> >> >> >
>> >> >> >> > ./configure --host=arm-linux-androideabi --prefix=$OUTPUT_DIR
>> >> >> >> > --with-zeromq=$OUTPUT_DIR CPPFLAGS="-fPIC
>> -I$OUTPUT_DIR/include"
>> >> >> >> > LDFLAGS="-L$OUTPUT_DIR/lib" --disable-version
>> >> >> >> >
>> >> >> >> > I get this error:
>> >> >> >> >
>> >> >> >> > "configure: error: cannot link with -lzmq"
>> >> >> >> >
>> >> >> >> > I have tried google for a few days now, but i cannot find any
>> >> >> solution
>> >> >> >> to
>> >> >> >> > this that works for me.
>> >> >> >> >
>> >> >> >> > Some comments on other solutions i have found, that did not
>> work
>> >> >> for
>> >> >> >> me:
>> >> >> >> >
>> >> >> >> > I have installed pretty much every compiler known to man, but
>> if
>> >> >> you
>> >> >> >> have
>> >> >> >> > any specific suggestions, please tell me.
>> >> >> >> > I have made no alterations to the source at all, simply
>> >> following
>> >> >> the
>> >> >> >> > instructions to the letter.
>> >> >> >> >
>> >> >> >> > Has someone else had this error? Is there something that is
>> >> >> >> "understood"
>> >> >> >> > in the guide, that i could be missing?
>> >> >> >> >
>> >> >> >> > Any help would be greatly appreciated!
>> >> >> >> >
>> >> >> >> > //Jonas Adler
>> >> >> >> >
>> >> >> >> > _______________________________________________
>> >> >> >> > zeromq-dev mailing list
>> >> >> >> > zeromq-dev at lists.zeromq.org
>> >> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >> >> > _______________________________________________
>> >> >> >> > zeromq-dev mailing list
>> >> >> >> > zeromq-dev at lists.zeromq.org
>> >> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >> >> >
>> >> >> >>
>> >> >> >>
>> >> >> >
>> >> >> >
>> >> >> > --
>> >> >> > Victor
>> >> >> > _______________________________________________
>> >> >> > zeromq-dev mailing list
>> >> >> > zeromq-dev at lists.zeromq.org
>> >> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >> >
>> >> >>
>> >> >>
>> >> >
>> >> >
>> >> > --
>> >> > Victor
>> >> > _______________________________________________
>> >> > zeromq-dev mailing list
>> >> > zeromq-dev at lists.zeromq.org
>> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >> >
>> >>
>> >>
>> >>
>> >
>> >
>> > --
>> > Victor
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
>
> --
> Victor
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



From sebastian.lauwers at gmail.com  Fri Jul 27 09:52:58 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Fri, 27 Jul 2012 09:52:58 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
	<CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
	<c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>
Message-ID: <501248DA.6000108@gmail.com>

On 27/07/12 09:43, jonas.adler at epiq.se wrote:
> I think i've found something. If I use sudo <configure> i get that
> arm-linux-androideabi-gcc isnt found, but when i dont use sudo and it is
> found, i get "C compiler cannot create executables"...

As a rule of thumb, you should never use the super user to perform the 
./configure && make steps of a compilation.

Only make install, when you are installing to a privileged prefix (/usr, 
or /usr/local), should be run as a privileged user.

-S.


From jonas.adler at epiq.se  Fri Jul 27 10:08:00 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Fri, 27 Jul 2012 10:08:00 +0200 (CEST)
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <501248DA.6000108@gmail.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
	<CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
	<c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>
	<501248DA.6000108@gmail.com>
Message-ID: <5235cc349efd283f1783f1e309269a8c.squirrel@mail.glesys.se>

Ok i'll think about that from here on, but if i don't use a privileged
user the ./conigure replies with "C compiler cannot create executables",
even though it find arm-linux-androideabi-gcc

> On 27/07/12 09:43, jonas.adler at epiq.se wrote:
>> I think i've found something. If I use sudo <configure> i get that
>> arm-linux-androideabi-gcc isnt found, but when i dont use sudo and it is
>> found, i get "C compiler cannot create executables"...
>
> As a rule of thumb, you should never use the super user to perform the
> ./configure && make steps of a compilation.
>
> Only make install, when you are installing to a privileged prefix (/usr,
> or /usr/local), should be run as a privileged user.
>
> -S.
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>



From victor at iso3103.net  Fri Jul 27 10:22:11 2012
From: victor at iso3103.net (Victor Perron)
Date: Fri, 27 Jul 2012 10:22:11 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <5235cc349efd283f1783f1e309269a8c.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
	<CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
	<c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>
	<501248DA.6000108@gmail.com>
	<5235cc349efd283f1783f1e309269a8c.squirrel@mail.glesys.se>
Message-ID: <CAF9FejFBdLtVG4zOu=RjA=NhVAtn=oA9UR1zawLP+7qkhDH9+Q@mail.gmail.com>

That's weird. You run the build process in /tmp, which I chose to minimize
those issues (it's generally chmodded 777) ; for the "C compiler cannot
create executables" issue as non-su,
- did you properly remove ALL the files from the previous compilation ?
even the directories, the cloned gits, the tar archive, etc ?
- if the issue persists, can you pastebin me your config.log again ?

On Fri, Jul 27, 2012 at 10:08 AM, <jonas.adler at epiq.se> wrote:

> Ok i'll think about that from here on, but if i don't use a privileged
> user the ./conigure replies with "C compiler cannot create executables",
> even though it find arm-linux-androideabi-gcc
>
> > On 27/07/12 09:43, jonas.adler at epiq.se wrote:
> >> I think i've found something. If I use sudo <configure> i get that
> >> arm-linux-androideabi-gcc isnt found, but when i dont use sudo and it is
> >> found, i get "C compiler cannot create executables"...
> >
> > As a rule of thumb, you should never use the super user to perform the
> > ./configure && make steps of a compilation.
> >
> > Only make install, when you are installing to a privileged prefix (/usr,
> > or /usr/local), should be run as a privileged user.
> >
> > -S.
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
> >
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>



-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/4f3c076b/attachment.htm 

From paolo.denti at gmail.com  Fri Jul 27 10:16:11 2012
From: paolo.denti at gmail.com (Paolo Denti)
Date: Fri, 27 Jul 2012 08:16:11 +0000 (UTC)
Subject: [zeromq-dev] iOS build
References: <AANLkTikzpyyZab4DzdREoVCEyi4SbFg3u2rBCrwxvm9c@mail.gmail.com>
Message-ID: <loom.20120727T101413-46@post.gmane.org>

Nick Forte <nick <at> clrtouch.com> writes:

> 
> 
> Has anyone successfully built libzmq.a and used it on iOS? 
> Before I spend time getting a 
> universal binary built for i386,arm6, and ...
> 
> 
> 
> 

Take a look at: http://paolodenti.blog.com/2012/07/24/zeromq-ios/

Cheers


From dave at directdataservices.co.uk  Fri Jul 27 12:42:54 2012
From: dave at directdataservices.co.uk (Dave Rutlidge)
Date: Fri, 27 Jul 2012 11:42:54 +0100
Subject: [zeromq-dev] C#/.NET CZMQ Library
Message-ID: <CAEGC5aK3eRjkp5z7Ap687JteiVpX_-jau3kG-uGza5qtmzYftg@mail.gmail.com>

Hi,

I'm working through The Guide and want to try the CZMQ examples but can't
find a library for .NET. How do I obtain it?

I've tried the usual places and got the sources from github but they won't
build in VS2012RC (hundreds on duplicate definition type errors).  Can
anyone help?

Thanks

Dave

-- 


Direct Data Services is a private limited company registered in England and Wales.
Registration number: 06064780.
Registered office:
Direct Data Services (DDS) Ltd
25 Grosvenor Road
Wrexham
LL11 1BT

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/9266d6e0/attachment.htm 

From jonas.adler at epiq.se  Fri Jul 27 13:04:54 2012
From: jonas.adler at epiq.se (jonas.adler at epiq.se)
Date: Fri, 27 Jul 2012 13:04:54 +0200 (CEST)
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <CAF9FejFBdLtVG4zOu=RjA=NhVAtn=oA9UR1zawLP+7qkhDH9+Q@mail.gmail.com>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
	<CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
	<c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>
	<501248DA.6000108@gmail.com>
	<5235cc349efd283f1783f1e309269a8c.squirrel@mail.glesys.se>
	<CAF9FejFBdLtVG4zOu=RjA=NhVAtn=oA9UR1zawLP+7qkhDH9+Q@mail.gmail.com>
Message-ID: <d3d8b060d7196fa1386ee07e933b8046.squirrel@mail.glesys.se>

I reinstalled Ubuntu (it's not my work machine, just using it to build for
android) and got a problem that the C compiler couldn't create executables
again but some searching on google i found that i needed "ia32-libs"
package because i was running a x64 ubuntu and the NDK is compiled for
32-bit.

After that 300mb package everything went just fine. I not have my zmq.jar
and libjzmq.so ! :)

So I can now confirm that it works (the compiling anyways).


But now I've come across another problem... I assume that android zmq is
identical to java, but when i tried an example (hwclient) I got a "FATAL
EXCEPTION: main, java.lang.NoClassDefFoundError: org.zeromq.ZMQ."
The only change i made to the code is "System.loadLibrary("jzmq");" at the
top.

the .jar is in /libs/ and has been added to build path and the .so is in
/libs/armeabi/

I've also added the internet permission to the android manifest.

> That's weird. You run the build process in /tmp, which I chose to minimize
> those issues (it's generally chmodded 777) ; for the "C compiler cannot
> create executables" issue as non-su,
> - did you properly remove ALL the files from the previous compilation ?
> even the directories, the cloned gits, the tar archive, etc ?
> - if the issue persists, can you pastebin me your config.log again ?
>
> On Fri, Jul 27, 2012 at 10:08 AM, <jonas.adler at epiq.se> wrote:
>
>> Ok i'll think about that from here on, but if i don't use a privileged
>> user the ./conigure replies with "C compiler cannot create executables",
>> even though it find arm-linux-androideabi-gcc
>>
>> > On 27/07/12 09:43, jonas.adler at epiq.se wrote:
>> >> I think i've found something. If I use sudo <configure> i get that
>> >> arm-linux-androideabi-gcc isnt found, but when i dont use sudo and it
>> is
>> >> found, i get "C compiler cannot create executables"...
>> >
>> > As a rule of thumb, you should never use the super user to perform the
>> > ./configure && make steps of a compilation.
>> >
>> > Only make install, when you are installing to a privileged prefix
>> (/usr,
>> > or /usr/local), should be run as a privileged user.
>> >
>> > -S.
>> >
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > zeromq-dev at lists.zeromq.org
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>> >
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
>
> --
> Victor
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>


From victor at iso3103.net  Fri Jul 27 14:34:37 2012
From: victor at iso3103.net (Victor Perron)
Date: Fri, 27 Jul 2012 14:34:37 +0200
Subject: [zeromq-dev] Re : Massive problems running ZMQ on android
In-Reply-To: <d3d8b060d7196fa1386ee07e933b8046.squirrel@mail.glesys.se>
References: <1343150764.17285.YahooMailClassic@web29502.mail.ird.yahoo.com>
	<df3cc44aa0526005b8c8072a8d752ab8.squirrel@mail.glesys.se>
	<CAF9FejFe+oKdepd5+M16q5PEhBZ-eva4beA8H-dwHc_XjiiVsA@mail.gmail.com>
	<29050c5d7b95b9b31443e332a8c255a1.squirrel@mail.glesys.se>
	<CAF9FejFt_CEs46mHHme5x_9aqFTKOXwHd5RS5ACHn1o9vutDUA@mail.gmail.com>
	<e6e754f5a642406b28116a814b5673ac.squirrel@mail.glesys.se>
	<CAF9FejHcizGOi+Gx9PgDhCTXyXxVicRTXtZq-akqzPTo=QPe0w@mail.gmail.com>
	<46694ac63d506ee97fec9c8076f8e9fa.squirrel@mail.glesys.se>
	<CAF9FejEk9wJtG5Xrvya0_WzaLoRj9qBL-n17X-Znai7kJLU11w@mail.gmail.com>
	<c31124f48c475f7ae025a3ad8aac0df2.squirrel@mail.glesys.se>
	<501248DA.6000108@gmail.com>
	<5235cc349efd283f1783f1e309269a8c.squirrel@mail.glesys.se>
	<CAF9FejFBdLtVG4zOu=RjA=NhVAtn=oA9UR1zawLP+7qkhDH9+Q@mail.gmail.com>
	<d3d8b060d7196fa1386ee07e933b8046.squirrel@mail.glesys.se>
Message-ID: <CAF9FejGUVbB7og7c1N12kkbKpp0bobyecO8=tJcP4rh4cTt7Tg@mail.gmail.com>

OK, that's some progress. I doubt you really need ia32-libs for anything
though, but whatever.

You're not supposed to add the System.loadLibrary() call by yourself as
ZMQ.java does it already.

And about the last issue, the ""FATAL EXCEPTION: main,
java.lang.NoClassDefFoundError: org.zeromq.ZMQ" makes me think that your
classpath is not *that* well set. But that is no more of my business, I
guess ;)

On Fri, Jul 27, 2012 at 1:04 PM, <jonas.adler at epiq.se> wrote:

> I reinstalled Ubuntu (it's not my work machine, just using it to build for
> android) and got a problem that the C compiler couldn't create executables
> again but some searching on google i found that i needed "ia32-libs"
> package because i was running a x64 ubuntu and the NDK is compiled for
> 32-bit.
>
> After that 300mb package everything went just fine. I not have my zmq.jar
> and libjzmq.so ! :)
>
> So I can now confirm that it works (the compiling anyways).
>
>
> But now I've come across another problem... I assume that android zmq is
> identical to java, but when i tried an example (hwclient) I got a "FATAL
> EXCEPTION: main, java.lang.NoClassDefFoundError: org.zeromq.ZMQ."
> The only change i made to the code is "System.loadLibrary("jzmq");" at the
> top.
>
> the .jar is in /libs/ and has been added to build path and the .so is in
> /libs/armeabi/
>
> I've also added the internet permission to the android manifest.
>
> > That's weird. You run the build process in /tmp, which I chose to
> minimize
> > those issues (it's generally chmodded 777) ; for the "C compiler cannot
> > create executables" issue as non-su,
> > - did you properly remove ALL the files from the previous compilation ?
> > even the directories, the cloned gits, the tar archive, etc ?
> > - if the issue persists, can you pastebin me your config.log again ?
> >
> > On Fri, Jul 27, 2012 at 10:08 AM, <jonas.adler at epiq.se> wrote:
> >
> >> Ok i'll think about that from here on, but if i don't use a privileged
> >> user the ./conigure replies with "C compiler cannot create executables",
> >> even though it find arm-linux-androideabi-gcc
> >>
> >> > On 27/07/12 09:43, jonas.adler at epiq.se wrote:
> >> >> I think i've found something. If I use sudo <configure> i get that
> >> >> arm-linux-androideabi-gcc isnt found, but when i dont use sudo and it
> >> is
> >> >> found, i get "C compiler cannot create executables"...
> >> >
> >> > As a rule of thumb, you should never use the super user to perform the
> >> > ./configure && make steps of a compilation.
> >> >
> >> > Only make install, when you are installing to a privileged prefix
> >> (/usr,
> >> > or /usr/local), should be run as a privileged user.
> >> >
> >> > -S.
> >> >
> >> > _______________________________________________
> >> > zeromq-dev mailing list
> >> > zeromq-dev at lists.zeromq.org
> >> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >> >
> >> >
> >>
> >>
> >> _______________________________________________
> >> zeromq-dev mailing list
> >> zeromq-dev at lists.zeromq.org
> >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >>
> >
> >
> >
> > --
> > Victor
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev at lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
>
>


-- 
Victor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/df62668b/attachment.htm 

From ian.barber at gmail.com  Fri Jul 27 17:10:06 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Fri, 27 Jul 2012 16:10:06 +0100
Subject: [zeromq-dev] [PATCH] [cppzmq] Add move constructor to message_t
In-Reply-To: <CADL5_siQi69HEojq8etz5QrbPtgzG_idDKS9Rba5Nvrd=dHX2Q@mail.gmail.com>
References: <OF21D5BAA1.FE45B344-ON80257A47.00446023-80257A47.0044CCF8@waters.com>
	<CADL5_siQi69HEojq8etz5QrbPtgzG_idDKS9Rba5Nvrd=dHX2Q@mail.gmail.com>
Message-ID: <CAEYnEwHOQmcaAAYN3TsMVQb+2LTqKwmtJfzVOBEVqtQB3Ep+0g@mail.gmail.com>

On Fri, Jul 27, 2012 at 8:03 AM, Pieter Hintjens <ph at imatix.com> wrote:
> Hi Richard,
>
> The process for contributions is described here:
> http://www.zeromq.org/docs:contributing
>
> Basically you want to make github pull requests against the cppzmq project.


Hi Pieter,

He both pull reqed and send a patch. Merged it in.

Ian

From linkfanel at yahoo.fr  Fri Jul 27 17:27:42 2012
From: linkfanel at yahoo.fr (Pierre Ynard)
Date: Fri, 27 Jul 2012 17:27:42 +0200
Subject: [zeromq-dev] ZMQ_MCAST_LOOP with PGM
In-Reply-To: <CAGFWLigj7NWGic6RgNfSkWLpZuhAM60E1rjgsCpLrKBCQuZU4w@mail.gmail.com>
Message-ID: <20120727152741.GA8276@via.ecp.fr>

> It can work well in a restricted environment, the disruptor team have
> shown great performance with Java and multicast loop but with the PGM
> protocol specifically there are problems and limitations that can
> break reliability.

Problems such as...?

I tried really hard but I couldn't figure out what you meant about the
unicast back channel and loopback, sorry.

-- 
Pierre Ynard
"Une ?me dans un corps, c'est comme un dessin sur une feuille de papier."

From diffuser78 at gmail.com  Fri Jul 27 19:26:13 2012
From: diffuser78 at gmail.com (diffuser78)
Date: Fri, 27 Jul 2012 10:26:13 -0700
Subject: [zeromq-dev] looking inside ZMQ queue
Message-ID: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>

Hi All,

I read the zguide and couldn't find the answer to it. I wanted to know if
there is a deterministic way to find out if a message that I sent has
actually reached the other end.

Let me explain the scenario in detail.

I am writing a lib (wrapped around ZMQ) for my app. App sends a message M
using my lib. The lib first persists the message on the disk or shared
memory, and then calls zmq_send() in non-blocking way. From my little
insight into ZMQ, I think if ZMQ can buffer the message in its queue
zmq_send() will return success and if it can't then it returns EAGAIN.

What I want to know is that, ZMQ might have returned success upon buffering
the message. But what if message is still in the queue and has not been
delivered to the receiver for some reasons (say slow receiver). Is there a
way I can peek into ZMQ queue to find out if my message M that I had sent
in a non blocking way was indeed sent on the wire successfully? My lib
needs to acknowledge to the app if there was an error after some time T for
it to timeout on the message M.

Please let me know your thoughts. I greatly appreciate your help.

Many thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/1ed3cce3/attachment.htm 

From ian.barber at gmail.com  Fri Jul 27 20:16:17 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Fri, 27 Jul 2012 19:16:17 +0100
Subject: [zeromq-dev] looking inside ZMQ queue
In-Reply-To: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
References: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
Message-ID: <CAEYnEwHYZoNFciiiU0SetR7HFR0ctaaJyEwWo7dGuVcAxndUnQ@mail.gmail.com>

On Fri, Jul 27, 2012 at 6:26 PM, diffuser78 <diffuser78 at gmail.com> wrote:
> What I want to know is that, ZMQ might have returned success upon buffering
> the message. But what if message is still in the queue and has not been
> delivered to the receiver for some reasons (say slow receiver). Is there a
> way I can peek into ZMQ queue to find out if my message M that I had sent in
> a non blocking way was indeed sent on the wire successfully? My lib needs to
> acknowledge to the app if there was an error after some time T for it to
> timeout on the message M.

It could be lost in the network, or the other side could crash while
processing. The best way is to send an acknowledgement back from the
receiver (also using ZeroMQ) and use that to determine whether the
other side has received and processed it.

Ian

From diffuser78 at gmail.com  Fri Jul 27 20:27:39 2012
From: diffuser78 at gmail.com (diffuser78)
Date: Fri, 27 Jul 2012 11:27:39 -0700
Subject: [zeromq-dev] looking inside ZMQ queue
In-Reply-To: <CAEYnEwHYZoNFciiiU0SetR7HFR0ctaaJyEwWo7dGuVcAxndUnQ@mail.gmail.com>
References: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
	<CAEYnEwHYZoNFciiiU0SetR7HFR0ctaaJyEwWo7dGuVcAxndUnQ@mail.gmail.com>
Message-ID: <CAKni0ef2aWzz9zzJ_-qSnV5cDMYcnX=OqLahSPhe0wk2YmQsNg@mail.gmail.com>

Sorry about my ignorance and laziness. I read this on the FAQ page:


*How do I determine how many messages are in queue?*

*This isn't possible. At any given time a message may be in the ?MQ sender
queue, the sender's kernel buffer, on the wire, in the receiver's kernel
buffer or in the receiver's ?MQ receiver queue. Furthermore, a ?MQ socket
can bind and/or connect to many peers. Each peer may have different
performance characteristics and therefore a different queue depth. Any
"queue depth" number is almost certainly wrong, so rather than provide
incorrect information the library avoids providing any view into this data.*
I had a related question but not same as above.

When ZMQ sent this message over TCP to the receivier and when receiver did
a FIN/ACK indicating that it is ACK'ing to the finish of the tranmission of
byte transfer, it indicates that the transport layer on the other end got
the message. Does ZMQ keep tab of this ? When does sender ZMQ removes this
message from its internal buffer. *I am only interested in knowing (in YES
or NO) whether this message is in sender zmq buffer or not*. Is that
possible ? If you can point me out, I can create a patch for it and send
for review.

Any thoughts are greatly appreciated.

DJ

On Fri, Jul 27, 2012 at 11:16 AM, Ian Barber <ian.barber at gmail.com> wrote:

> On Fri, Jul 27, 2012 at 6:26 PM, diffuser78 <diffuser78 at gmail.com> wrote:
> > What I want to know is that, ZMQ might have returned success upon
> buffering
> > the message. But what if message is still in the queue and has not been
> > delivered to the receiver for some reasons (say slow receiver). Is there
> a
> > way I can peek into ZMQ queue to find out if my message M that I had
> sent in
> > a non blocking way was indeed sent on the wire successfully? My lib
> needs to
> > acknowledge to the app if there was an error after some time T for it to
> > timeout on the message M.
>
> It could be lost in the network, or the other side could crash while
> processing. The best way is to send an acknowledgement back from the
> receiver (also using ZeroMQ) and use that to determine whether the
> other side has received and processed it.
>
> Ian
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/42a4de4b/attachment.htm 

From stu at compuserve.com  Fri Jul 27 21:23:35 2012
From: stu at compuserve.com (Stuart Brandt)
Date: Fri, 27 Jul 2012 15:23:35 -0400
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
Message-ID: <5012EAB7.80500@compuserve.com>

I get an assertion during a call to zmq_recv on 2 different OSes (OSX 
10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and 
a recent build off latest at github.com/zeromq/libzmq).

The assertion is:
Assertion failed: (msg_->flags () & msg_t::identity) == 0 (router.cpp:220)
Aborted

The setup involves:
1) the simple message queue broker found at 
https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
I modified the #include to look for zmq.h rather than zhelpers.h to get 
it to compile with ZMQ 3.x

2) a minimal server implementation that connects to the broker and does 
a zmq_recv. Code is:
//
//  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
//  Connects to tcp://localhost:5560
//
#include <stdio.h>
#include <zmq.h>

int main (void)
{
     char buf[80];   // buffer for zmq_recv call
     int len;        // len returned by zmq_recv
     void *ctx;      // ZMQ context
     void *sock;     // ZMQ socket

     // init ZMQ context, create a REP socket, and connect to ROUTER
     ctx = zmq_ctx_new();
     sock = zmq_socket(ctx,ZMQ_REP);
     zmq_connect (sock, "tcp://localhost:5560");

     // do a basic zmq_recv call...warn if it fails
     len = zmq_recv(sock,buf,sizeof(buf),0);
     if (len < 0) {
         printf("zmq_recv failed - %s\n",zmq_strerror(errno));
     }

     // Done. Clean up and end
     zmq_close (sock);
     zmq_term (ctx);
     return 0;
}


Steps to repro are start server, start broker, kill broker (spec. 
ctrl-C), restart broker. Server aborts with the above assertion.

Is there something my code should be doing differently?

Thanks....Stuart




From bklooste at gmail.com  Sat Jul 28 00:46:02 2012
From: bklooste at gmail.com (Bennie Kloosteman)
Date: Sat, 28 Jul 2012 06:46:02 +0800
Subject: [zeromq-dev] looking inside ZMQ queue
In-Reply-To: <CAKni0ef2aWzz9zzJ_-qSnV5cDMYcnX=OqLahSPhe0wk2YmQsNg@mail.gmail.com>
References: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
	<CAEYnEwHYZoNFciiiU0SetR7HFR0ctaaJyEwWo7dGuVcAxndUnQ@mail.gmail.com>
	<CAKni0ef2aWzz9zzJ_-qSnV5cDMYcnX=OqLahSPhe0wk2YmQsNg@mail.gmail.com>
Message-ID: <CAJT18ia4C+2HVOECzwJCmcyRu6huEFB0GvcqEPx=Z352iu5c6A@mail.gmail.com>

Like most systems ZeroMQ will just send the message to the transport when
the transport has received the message ( eg TCP/IP )it is removed from
ZeroMQ. There are no guarantees it has even been sent , yet alone that the
transport layer on the other side has the message.  This is pretty standard
TCP/IP provides no guarantees the other service has received the message (
only the receiver TCp/IP stack) so you need to code this in your
application logic.

Also if you are persisting you do not need a light weight system ,
persistence is always slow and heavy , killing any performance ( Normally
non persistent messaging systems do at least 10* the messages / second to
persistent ones) . For persistence I would use a queue like MSMQ or the IBM
equivalent   . Queues also help in knowing if the other side has received
it.

Ben



On Sat, Jul 28, 2012 at 2:27 AM, diffuser78 <diffuser78 at gmail.com> wrote:

> Sorry about my ignorance and laziness. I read this on the FAQ page:
>
>
> *How do I determine how many messages are in queue?*
>
> *This isn't possible. At any given time a message may be in the ?MQ
> sender queue, the sender's kernel buffer, on the wire, in the receiver's
> kernel buffer or in the receiver's ?MQ receiver queue. Furthermore, a ?MQ
> socket can bind and/or connect to many peers. Each peer may have different
> performance characteristics and therefore a different queue depth. Any
> "queue depth" number is almost certainly wrong, so rather than provide
> incorrect information the library avoids providing any view into this data.
> *
> I had a related question but not same as above.
>
> When ZMQ sent this message over TCP to the receivier and when receiver did
> a FIN/ACK indicating that it is ACK'ing to the finish of the tranmission of
> byte transfer, it indicates that the transport layer on the other end got
> the message. Does ZMQ keep tab of this ? When does sender ZMQ removes this
> message from its internal buffer. *I am only interested in knowing (in
> YES or NO) whether this message is in sender zmq buffer or not*. Is that
> possible ? If you can point me out, I can create a patch for it and send
> for review.
>
> Any thoughts are greatly appreciated.
>
> DJ
>
> On Fri, Jul 27, 2012 at 11:16 AM, Ian Barber <ian.barber at gmail.com> wrote:
>
>> On Fri, Jul 27, 2012 at 6:26 PM, diffuser78 <diffuser78 at gmail.com> wrote:
>> > What I want to know is that, ZMQ might have returned success upon
>> buffering
>> > the message. But what if message is still in the queue and has not been
>> > delivered to the receiver for some reasons (say slow receiver). Is
>> there a
>> > way I can peek into ZMQ queue to find out if my message M that I had
>> sent in
>> > a non blocking way was indeed sent on the wire successfully? My lib
>> needs to
>> > acknowledge to the app if there was an error after some time T for it to
>> > timeout on the message M.
>>
>> It could be lost in the network, or the other side could crash while
>> processing. The best way is to send an acknowledgement back from the
>> receiver (also using ZeroMQ) and use that to determine whether the
>> other side has received and processed it.
>>
>> Ian
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120728/047f3c31/attachment.htm 

From steven.mccoy at miru.hk  Sat Jul 28 01:03:12 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Fri, 27 Jul 2012 19:03:12 -0400
Subject: [zeromq-dev] ZMQ_MCAST_LOOP with PGM
In-Reply-To: <20120727152741.GA8276@via.ecp.fr>
References: <CAGFWLigj7NWGic6RgNfSkWLpZuhAM60E1rjgsCpLrKBCQuZU4w@mail.gmail.com>
	<20120727152741.GA8276@via.ecp.fr>
Message-ID: <CAGFWLiidax==0gKR8uXdc8Uie8FMcX-koNcpAmz_p7v1x1LdeA@mail.gmail.com>

On 27 July 2012 11:27, Pierre Ynard <linkfanel at yahoo.fr> wrote:

> > It can work well in a restricted environment, the disruptor team have
> > shown great performance with Java and multicast loop but with the PGM
> > protocol specifically there are problems and limitations that can
> > break reliability.
>
> Problems such as...?
>

The OS only forwards packets to the first open socket.


> I tried really hard but I couldn't figure out what you meant about the
> unicast back channel and loopback, sorry.
>
>
Check the RFC's on PGM and PGMCC for details of the protocols in question,

http://www.ietf.org/rfc/rfc3208.txt
http://www.ietf.org/proceedings/57/I-D/draft-ietf-rmt-bb-pgmcc-02.txt

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120727/202f9ea6/attachment.htm 

From ph at imatix.com  Sat Jul 28 05:33:58 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Sat, 28 Jul 2012 12:33:58 +0900
Subject: [zeromq-dev] looking inside ZMQ queue
In-Reply-To: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
References: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
Message-ID: <CADL5_shDFYhFPtB0-uS=Yfxzqa0dsMNtFgr8SajtP4t6X19pEg@mail.gmail.com>

On Sat, Jul 28, 2012 at 2:26 AM, diffuser78 <diffuser78 at gmail.com> wrote:

> I read the zguide and couldn't find the answer to it. I wanted to know if
> there is a deterministic way to find out if a message that I sent has
> actually reached the other end.

The short answer is that if you need information from the receiver,
you MUST send that back as a message.

The longer answer is in the Guide, in the section on request-reply
reliability (Chapter 4).

-Pieter

From oj at buffered.io  Sat Jul 28 08:57:19 2012
From: oj at buffered.io (OJ Reeves)
Date: Sat, 28 Jul 2012 16:57:19 +1000
Subject: [zeromq-dev] C#/.NET CZMQ Library
In-Reply-To: <CAEGC5aK3eRjkp5z7Ap687JteiVpX_-jau3kG-uGza5qtmzYftg@mail.gmail.com>
References: <CAEGC5aK3eRjkp5z7Ap687JteiVpX_-jau3kG-uGza5qtmzYftg@mail.gmail.com>
Message-ID: <CAG63A8s51cWwRne_VNMpEpmMbbDtSiaL9qyh06uD62D2ojN7aQ@mail.gmail.com>

Dave,

Do you absolutely need to build it from source? Why not use the NuGet
packages?

http://nuget.org/packages/clrzmq
http://nuget.org/packages/clrzmq-x64

Cheers.
OJ



On Fri, Jul 27, 2012 at 8:42 PM, Dave Rutlidge <
dave at directdataservices.co.uk> wrote:

> Hi,
>
> I'm working through The Guide and want to try the CZMQ examples but can't
> find a library for .NET. How do I obtain it?
>
> I've tried the usual places and got the sources from github but they won't
> build in VS2012RC (hundreds on duplicate definition type errors).  Can
> anyone help?
>
> Thanks
>
> Dave
>
> Direct Data Services is a private limited company registered in England and Wales.
> Registration number: 06064780.
> Registered office:
> Direct Data Services (DDS) Ltd
> 25 Grosvenor Road
> Wrexham
> LL11 1BT
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>


-- 

OJ Reeves
+61 431 952 586
http://buffered.io/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120728/972f1b3e/attachment.htm 

From breese at mail1.stofanet.dk  Sat Jul 28 12:52:01 2012
From: breese at mail1.stofanet.dk (Bjorn Reese)
Date: Sat, 28 Jul 2012 12:52:01 +0200
Subject: [zeromq-dev] Router to router delay
Message-ID: <5013C451.1080000@mail1.stofanet.dk>

I am writing a peer-to-peer system on top of a router-to-router
connection. As there can be multiple peers, I set their identity.

It appears that router pairs have to exchange their identity before
they can exchange messages. Any message I send before the identity
is exhanged, are silently discarded.

The only solution I have seen to this problem, is to sleep for one
second before sending the first message. Unfortunately, there is no
guarantee that one second is sufficient. A slow connection may take
longer than one second to exchange the identity. So this is not a
feasible solution.

I see three alternatives:

1. Let ZeroMQ queue the messages and send them once the identity has
    been exchanged. This is my preferred solution seen from a users
    point of view.

2. Let zmq_send() return EAGAIN until the identity has been exchanged.

3. Add an event to notify the application that the identity has been
    exchanged (e.g. via the new zmq_ctx_monitor_event() functionality.)

From cfkaran2 at gmail.com  Sat Jul 28 13:02:02 2012
From: cfkaran2 at gmail.com (Cem Karan)
Date: Sat, 28 Jul 2012 07:02:02 -0400
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <5013C451.1080000@mail1.stofanet.dk>
References: <5013C451.1080000@mail1.stofanet.dk>
Message-ID: <91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>

Can you describe your system a bit more?  Are the peers on the same LAN, or do they need to peer across the internet?  Do the peers already know about each other (as in, a small group with well-known identities), or do they have to discover each other?  etc...

Thanks,
Cem Karan

On Jul 28, 2012, at 6:52 AM, Bjorn Reese wrote:

> I am writing a peer-to-peer system on top of a router-to-router
> connection. As there can be multiple peers, I set their identity.
> 
> It appears that router pairs have to exchange their identity before
> they can exchange messages. Any message I send before the identity
> is exhanged, are silently discarded.
> 
> The only solution I have seen to this problem, is to sleep for one
> second before sending the first message. Unfortunately, there is no
> guarantee that one second is sufficient. A slow connection may take
> longer than one second to exchange the identity. So this is not a
> feasible solution.
> 
> I see three alternatives:
> 
> 1. Let ZeroMQ queue the messages and send them once the identity has
>    been exchanged. This is my preferred solution seen from a users
>    point of view.
> 
> 2. Let zmq_send() return EAGAIN until the identity has been exchanged.
> 
> 3. Add an event to notify the application that the identity has been
>    exchanged (e.g. via the new zmq_ctx_monitor_event() functionality.)
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1581 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120728/94b9d217/attachment.bin 

From breese at mail1.stofanet.dk  Sat Jul 28 14:11:47 2012
From: breese at mail1.stofanet.dk (Bjorn Reese)
Date: Sat, 28 Jul 2012 14:11:47 +0200
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
References: <5013C451.1080000@mail1.stofanet.dk>
	<91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
Message-ID: <5013D703.60502@mail1.stofanet.dk>

On 2012-07-28 13:02, Cem Karan wrote:
> Can you describe your system a bit more?  Are the peers on the same LAN, or do they need to peer across the internet?  Do the peers already know about each other (as in, a small group with well-known identities), or do they have to discover each other?  etc...

Initially, all peers are on the same LAN where they discover each other
via ZeroConf (Bonjour). These local peers constitutes one cluster.
Later this has to be extended to several clusters communicating over
WAN. The exact means for this has not been determined yet.

From ivan at blackpx.com  Sat Jul 28 16:34:15 2012
From: ivan at blackpx.com (Ivan)
Date: Sat, 28 Jul 2012 14:34:15 +0000 (UTC)
Subject: [zeromq-dev] If not Majordomo for asynchronous workers,
	then what?
References: <CAEiCcQBXNrGyWqj17tgTDzJw4gVqmHp7Vj-QoJYj4zNYnaMgow@mail.gmail.com>
	<CADL5_sg5qBPbqZaQVBS_Z_WD_fN2WwvtJbxOxaNGhSaTcGJ5kg@mail.gmail.com>
Message-ID: <loom.20120728T162448-153@post.gmane.org>

Pieter,

Pieter Hintjens <ph <at> imatix.com> writes:

> 
> Hi Ivan,
> 
> Majordomo is based on a 1-to-1 request-reply model for workers, yes.
> You should be able to quite easily modify MDP to allow multiple
> replies, perhaps by adding a "there's more coming" indicator to
> replies that the broker uses to route several replies back to the same
> client.

It seems like I will have to map a clientOrderID to a particular client. That 
way, when the FIX engine sends back more messages, I can look up this order ID, 
get the client, and send it back to that particular client.

> Modifying MDP and the implementations will be simpler than writing
> your own from scratch.

Right. I believe what I describe above is the correct modification to the 
Broker. I have to understand the magic by which the broker knows how to route a 
message back to the client. It would be a great addition to MDP if I could call 
a function that, given a message, would break the multipart message for me. This 
way I could hash the route to the client based on some key, in my case, OrderID.

> You can also use other architectures; MDP is specifically aimed at
> allowing multiple redundant workers per service type. If you have a
> single well-known gateway per exchange you can just connect each
> client to each exchange gateway directly and work fully
> asynchronously.

Having the clients know of the routes would break encapsulation, to say nothing 
of the fact that there can be all sort of different protocols like FIX, ITCH, 
proprietary binary, etc. The idea is that a client can send normalized/generic 
order to a broker, who then fills in the appropriate details as to how to get 
that order to a particular exchange. On the way back from the exchange, the 
reverse process occurs, where a given message from an exchange is normalized 
back into a simpler internal message. So complexity is hidden from the client.


> -Pieter

Thanks again Peiter for your suggestions. Is there any way I can contact you 
directly?

Ivan


> On Thu, Jul 26, 2012 at 10:29 PM, Ivan Figueredo <ivan <at> blackpx.com> 
wrote:
> > I am writing a distributed trading manager. There are many different
> > liquidity providers, and the client would like to route to a given exchange.
> > With this in mind, I thought that using the Majordomo pattern with a broker
> > in between clients and workers asking a particular [named] worker to do its
> > work. However at least as I understand it, Majordomo won't work in this
> > scenario.
> >
> > Let's take the FIX protocol. Since FIX is asynchronous, I believe I can't
> > use the Majordomo. You may send an order to a FIX engine, and that one order
> > may send back many different messages. For example, that one order may see
> > the following responses: Insert, Pending, New, Partial Fill, then finally
> > Fill.
> >
> > In the Majordomo pattern, a section of code in the worker looks like this
> >
> > mdwrk session ("tcp://localhost:5555", sourceStr.c_str(), verbose);
> >
> > zmsg *reply = 0;
> > while (1) {
> >     zmsg *request = session.recv (reply);
> >     if (request == 0) {
> >         break;              //  Worker was interrupted
> >     }
> >
> >     //reply = request;        //  Echo is complex... 
> >     reply = new zmsg(sourceStr.c_str());
> > }
> >
> >
> > This assumes one request, and one reply. Since my application is not
> > strictly a single request, that gets a single response as in the code above,
> > do I have to drop Majordomo ?
> >
> >
> > Thanks!
> >
> >
> > _______________________________________________
> > zeromq-dev mailing list
> > zeromq-dev <at> lists.zeromq.org
> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> >
> 




From sebastian.lauwers at gmail.com  Sat Jul 28 18:05:28 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Sat, 28 Jul 2012 18:05:28 +0200
Subject: [zeromq-dev] Question about constness
In-Reply-To: <50123BA8.3080806@gmail.com>
References: <500E28F6.1040608@gmail.com>
	<CAESgSZqOZGD+oNrzSuNF_JjH8sbqW0aZ1--VVu67=r=Y=scG1g@mail.gmail.com>
	<CADL5_sivSuCgWDJ223ue0uNsRsgV67=KZe6kY+o5fmv-wxkTaA@mail.gmail.com>
	<50123BA8.3080806@gmail.com>
Message-ID: <50140DC8.1020708@gmail.com>

On 07/27/2012 08:56 AM, Sebastian Lauwers wrote:
> Very well, I'll try to provide a patch tomorrow.

I guess I hadn't looked at the CPP API closely before answering to Ming, 
or offering a patch. I'm not seeing anything in that API that doesn't 
have proper constness applied.

Apologies to Martin for blindly assuming it wasn't done.

-S.



From cfkaran2 at gmail.com  Sat Jul 28 20:48:08 2012
From: cfkaran2 at gmail.com (Cem Karan)
Date: Sat, 28 Jul 2012 14:48:08 -0400
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <5013D703.60502@mail1.stofanet.dk>
References: <5013C451.1080000@mail1.stofanet.dk>
	<91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
	<5013D703.60502@mail1.stofanet.dk>
Message-ID: <D9AD4F80-1DA3-4E96-9002-05F184FACC00@gmail.com>

OK, you're doing something similar to what I'm trying to do, except that you're planning on expanding to a WAN.  I'm going to outline what I'm doing, but I want to note that what I have are a number of pieces of test code that let me make sure that the way I'm using zmq makes sense.  I haven't tried to push large amounts of data through yet, so it may break.  With that disclaimer in mind...

I started out with router-router connections, but switched to pub-sub over epgm and administratively scoped multicast addresses (http://en.wikipedia.org/wiki/Multicast_address#Administratively_Scoped_IPv4_Multicast_addresses) for precisely the reasons you outlined in your original email; it isn't easy distributing the addresses.  In my code, each node has one publisher socket and one subscriber socket.  All publishers and all subscribers are connected to the same address and port.   This means that I effectively have a reliable broadcast channel (epgm ensures that all receivers get the message, in the correct order, etc.).  I wrote a simple control layer/protocol on top that allows all nodes to communicate some minimal control information, which in your case might include zmq URIs.  That should allow router-router connections.

I don't know if the IANA has any addresses in the ad-hoc block (http://en.wikipedia.org/wiki/Multicast_address#AD-HOC_block) which are open to individual developers, but that is another thing to look at; if you can get permission to use an address in there, then the pub-sub trick would work across the internet. 

Thanks,
Cem Karan

On Jul 28, 2012, at 8:11 AM, Bjorn Reese wrote:

> On 2012-07-28 13:02, Cem Karan wrote:
>> Can you describe your system a bit more?  Are the peers on the same LAN, or do they need to peer across the internet?  Do the peers already know about each other (as in, a small group with well-known identities), or do they have to discover each other?  etc...
> 
> Initially, all peers are on the same LAN where they discover each other
> via ZeroConf (Bonjour). These local peers constitutes one cluster.
> Later this has to be extended to several clusters communicating over
> WAN. The exact means for this has not been determined yet.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1581 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120728/5647a062/attachment.bin 

From artem.vysochyn at gmail.com  Sun Jul 29 00:39:52 2012
From: artem.vysochyn at gmail.com (Artem V12)
Date: Sun, 29 Jul 2012 01:39:52 +0300
Subject: [zeromq-dev] Pls. advice on design/architecture.
Message-ID: <CAN5_7Dw-rkoqWbZ5ahuvEH7OkvVR9jeKiMdtF5Zj59k2-nzhDw@mail.gmail.com>

Hello there, 0Community!

I want to build some cool notification system based on 0MQ, but this's
gonna be challening effort. I read the guide(almost all) but didn't find
what can help me. So here's the thing:
I have a persistent layer where I store users and their preferences. The
preferences, essentially, are the filters against neverending flow of some
events (at this point it's doesn't matter what events represent by itself)
and can be updated at any time by user. One fact about events and user'
preferences - they can be matched together by certain prefix. So, at the
end of the day I have to have notification system which on arrival of
certain event has to match (by prefix in the messsage) corresponding user'
pref, then check that event matched is 'good', and send email to user.
Essentially I have to track the updates of certain stuff.

So I after reading the guide, it's more/less clear to me how to impl. this.
1. On one side there will be fanout broadcast of events which will bombard
by messages the two 0MQ boxes (second is backup).
2. On the other side - clients which has to be subscribed (by user' pref)
onto one of the two endpoints.

If understood 0MQ idea correctly then making some component to listen for
some events isn't hard thing. But here's the issue: user' prefs (i.e.
filters) are persistent and, worse, are subject to Create/Update/Delete.
So, say, in jvm, I will pull a user' pref, make it listen on ZMQ.SUB
socket, but while I was doing this user could Update/Delete his pref. Or
better - once user' pref is created - let it listen 'from the birth', but
the first restart of the box where user' pref is listening for events will
end the connection.

In general the problem I faced is: how to make persistent things to listen
in reliable manner?


Thanks in advance!

-regards
-artemv
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120729/f99baa56/attachment.htm 

From ph at imatix.com  Sun Jul 29 05:13:27 2012
From: ph at imatix.com (Pieter Hintjens)
Date: Sun, 29 Jul 2012 12:13:27 +0900
Subject: [zeromq-dev] If not Majordomo for asynchronous workers,
	then what?
In-Reply-To: <loom.20120728T162448-153@post.gmane.org>
References: <CAEiCcQBXNrGyWqj17tgTDzJw4gVqmHp7Vj-QoJYj4zNYnaMgow@mail.gmail.com>
	<CADL5_sg5qBPbqZaQVBS_Z_WD_fN2WwvtJbxOxaNGhSaTcGJ5kg@mail.gmail.com>
	<loom.20120728T162448-153@post.gmane.org>
Message-ID: <CADL5_shp--yXqgjAq05VTogbgOEWkK4vpdq0sOZLKR9_EHK=WA@mail.gmail.com>

On Sat, Jul 28, 2012 at 11:34 PM, Ivan <ivan at blackpx.com> wrote:

> Right. I believe what I describe above is the correct modification to the
> Broker. I have to understand the magic by which the broker knows how to route a
> message back to the client. It would be a great addition to MDP if I could call
> a function that, given a message, would break the multipart message for me. This
> way I could hash the route to the client based on some key, in my case, OrderID.

It's explained in http://rfc.zeromq.org/spec:7 and in the Guide that
develops MDP.

-Pieter

From paul at colomiets.name  Sun Jul 29 16:49:01 2012
From: paul at colomiets.name (Paul Colomiets)
Date: Sun, 29 Jul 2012 17:49:01 +0300
Subject: [zeromq-dev] If not Majordomo for asynchronous workers,
	then what?
In-Reply-To: <loom.20120728T162448-153@post.gmane.org>
References: <CAEiCcQBXNrGyWqj17tgTDzJw4gVqmHp7Vj-QoJYj4zNYnaMgow@mail.gmail.com>
	<CADL5_sg5qBPbqZaQVBS_Z_WD_fN2WwvtJbxOxaNGhSaTcGJ5kg@mail.gmail.com>
	<loom.20120728T162448-153@post.gmane.org>
Message-ID: <CAA0gF6oJH_nOHbgMVfof3MTSxUmqovbv0nNTYUOq3nks=L80Gw@mail.gmail.com>

Hi Ivan,

On Sat, Jul 28, 2012 at 5:34 PM, Ivan <ivan at blackpx.com> wrote:
>> Modifying MDP and the implementations will be simpler than writing
>> your own from scratch.
>
> Right. I believe what I describe above is the correct modification to the
> Broker. I have to understand the magic by which the broker knows how to route a
> message back to the client. It would be a great addition to MDP if I could call
> a function that, given a message, would break the multipart message for me. This
> way I could hash the route to the client based on some key, in my case, OrderID.
>

There are actually two schools here:

First, proposes to modify the broker, as you described.

Second, requires to decompose the communication into patterns. In your
case it's as follows:

1. You have two connections: requrest/reply, and pub/sub
2. Before doing request, you create and OrderID, and you subscribe on
that OrderID in pub/sub
3. Than you do the request. There is no reply until the whole order is processed
4. The order status updates are sent by pub/sub
5. When everything is done, you receive final reply on request

I usually choose the latter approach, for the following reasons:

1. The solution consists of two clearly defined patterns
2. It leverages zeromq ability to introduce intermediate nodes (as it
uses bare patterns)
3. It allows to run brokerless when setup is small (e.g. for development)
4. It allows to use simple request-reply for clients which do not need
intermediate statuses
5. It allows another peer/application to follow intermediate statuses
(so gives more flexibility to the client)


-- 
Paul

From arkrost at gmail.com  Sun Jul 29 12:10:16 2012
From: arkrost at gmail.com (Arkady Rost)
Date: Sun, 29 Jul 2012 14:10:16 +0400
Subject: [zeromq-dev] zeromq 3.2 tests fails
Message-ID: <CAP8nFi7gCbdP9kTp-d6h4rYg7VYmfX4vR76ZgjeZ=b4aSvzhBA@mail.gmail.com>

Hi! I've just installed zeromq 3.2 from sources but 2 tests fails. They are
test_monitor and test_shutdown_stress. Here is full check log:
https://gist.github.com/3197194. I think that the problem with test_monitor
is related with issue #399. Any ideas what should I fix?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120729/ff433372/attachment.htm 

From sebastian.lauwers at gmail.com  Mon Jul 30 08:44:33 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Mon, 30 Jul 2012 08:44:33 +0200
Subject: [zeromq-dev] zeromq 3.2 tests fails
In-Reply-To: <CAP8nFi7gCbdP9kTp-d6h4rYg7VYmfX4vR76ZgjeZ=b4aSvzhBA@mail.gmail.com>
References: <CAP8nFi7gCbdP9kTp-d6h4rYg7VYmfX4vR76ZgjeZ=b4aSvzhBA@mail.gmail.com>
Message-ID: <CAESgSZpyoA9gAsjh83oD0BPCyG6EY1kcm8TVXtsFpzd=Ti8mnQ@mail.gmail.com>

On 29 July 2012 12:10, Arkady Rost <arkrost at gmail.com> wrote:
> Hi!

Hi Arkady,

> I've just installed zeromq 3.2 from sources but 2 tests fails. They are
> test_monitor and test_shutdown_stress.

Could you confirm which repository/branch you are using? From the
gist, I'd say it's the zeromq 3.x repository
(https://github.com/zeromq/zeromq3-x.git). Also, could you indicate at
what position your HEAD is?

I've not been able to reproduce the segfault problem using the latest
HEAD (d3518eafe4b58c0d7606ef0d9035df3ec52c22d4), however, the
test_monitor does indeed fail.

git bisect shows that the regression might be introduced by commit
38a0710b89b7bfe81f7250c5e9a96e024cfabe62 (Fixed issue #399
socket_monitor test, fromXMbIPEK <welcome.jet at gmail.com>). However, I
don't think this commit is the actual origin.


From Richard_Newton at waters.com  Mon Jul 30 09:55:26 2012
From: Richard_Newton at waters.com (Richard_Newton at waters.com)
Date: Mon, 30 Jul 2012 08:55:26 +0100
Subject: [zeromq-dev] [PATCH] [cppzmq] Add move constructor to message_t
In-Reply-To: <CAEYnEwHOQmcaAAYN3TsMVQb+2LTqKwmtJfzVOBEVqtQB3Ep+0g@mail.gmail.com>
References: <OF21D5BAA1.FE45B344-ON80257A47.00446023-80257A47.0044CCF8@waters.com>
	<CADL5_siQi69HEojq8etz5QrbPtgzG_idDKS9Rba5Nvrd=dHX2Q@mail.gmail.com>
	<CAEYnEwHOQmcaAAYN3TsMVQb+2LTqKwmtJfzVOBEVqtQB3Ep+0g@mail.gmail.com>
Message-ID: <OF41A37A04.111F3D40-ON80257A4B.002AF330-80257A4B.002B8727@waters.com>


Hi Ian,

Thanks, but merged it in where?  I cannot see the changes at https://github.com/zeromq/cppzmq.

So I am clear, what is the protocol for submitting patches?  Should I be sending the patch to the list for review as well as doing a pull request or
do you just want the pull request?

Sending the patch to the mailing list as well would seem to be sensible so it can be reviewed, I notice there is some talk on the list about
constness, I have another pull request that adds const data access to message_t that probably isn't that visible to people on the list.

Regards,

Richard.

  Richard Newton                         Waters Logo
  Principal Software
  Engineer



  [T]?+44 (0) 161 946                    Waters
  2685                                   Corporation
  [F]?+44 (0) 161 946                    Floats Road
  2480                                   Wythenshawe
  [W]?www.waters.com                     Manchester
  [E]                                    M23 9LZ
  richard_newton at waters.com              UK






Micromass UK Limited is registered in England under No 3162904
Registered office: Floats Road, Wythenshawe, Manchester M23 9LZ




From:	"Ian Barber" <ian.barber at gmail.com>
To:	"ZeroMQ development list" <zeromq-dev at lists.zeromq.org>
Date:	27/07/2012 04:10 PM
Subject:	Re: [zeromq-dev] [PATCH] [cppzmq] Add move constructor to message_t
Sent by:	zeromq-dev-bounces at lists.zeromq.org



On Fri, Jul 27, 2012 at 8:03 AM, Pieter Hintjens <ph at imatix.com> wrote:
> Hi Richard,
>
> The process for contributions is described here:
> http://www.zeromq.org/docs:contributing
>
> Basically you want to make github pull requests against the cppzmq project.


Hi Pieter,

He both pull reqed and send a patch. Merged it in.

Ian
_______________________________________________
zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


===========================================================
The information in this email is confidential, and is intended solely for the addressee(s). 
Access to this email by anyone else is unauthorized and therefore prohibited.  If you are 
not the intended recipient you are notified that disclosing, copying, distributing or taking 
any action in reliance on the contents of this information is strictly prohibited and may be unlawful.
===========================================================
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/b6630a08/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 70263135.gif
Type: image/gif
Size: 3839 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/b6630a08/attachment.gif 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ecblank.gif
Type: image/gif
Size: 45 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/b6630a08/attachment-0001.gif 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: graycol.gif
Type: image/gif
Size: 105 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/b6630a08/attachment-0002.gif 

From breese at mail1.stofanet.dk  Mon Jul 30 10:33:47 2012
From: breese at mail1.stofanet.dk (Bjorn Reese)
Date: Mon, 30 Jul 2012 10:33:47 +0200
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <D9AD4F80-1DA3-4E96-9002-05F184FACC00@gmail.com>
References: <5013C451.1080000@mail1.stofanet.dk>
	<91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
	<5013D703.60502@mail1.stofanet.dk>
	<D9AD4F80-1DA3-4E96-9002-05F184FACC00@gmail.com>
Message-ID: <501646EB.9070807@mail1.stofanet.dk>

We are straying off-topic.

I also use publish-subscribe in my system, but I need router-router for
point-to-point interaction.

So my original problem about messages being silently discarded before
the identity exchange is finalized is still relevant.

Let me elaborate on the original problem with an example. The example
is against https://github.com/zeromq/zeromq3-x and error handling
has been removed for brevity (but it is present in my original code)

   void *client = zmq_socket(context, ZMQ_ROUTER);
   zmq_setsockopt(client, ZMQ_IDENTITY, identity, strlen(identity));
   zmq_connect(client, endpoint);

   sleep(1); /* Wait for identity exchange to finish */

   zmq_msg_t message;
   zmq_msg_init_size(&message, strlen(data));
   memcpy(zmq_msg_data(&message), data, strlen(data));
   int rc = zmq_sendmsg(client, &message, 0);
   assert(rc == 0);

If the sleep() function above is removed, then the subsequent message
is silently ignored.


On 2012-07-28 20:48, Cem Karan wrote:
> OK, you're doing something similar to what I'm trying to do, except that you're planning on expanding to a WAN.  I'm going to outline what I'm doing, but I want to note that what I have are a number of pieces of test code that let me make sure that the way I'm using zmq makes sense.  I haven't tried to push large amounts of data through yet, so it may break.  With that disclaimer in mind...
>
> I started out with router-router connections, but switched to pub-sub over epgm and administratively scoped multicast addresses (http://en.wikipedia.org/wiki/Multicast_address#Administratively_Scoped_IPv4_Multicast_addresses) for precisely the reasons you outlined in your original email; it isn't easy distributing the addresses.  In my code, each node has one publisher socket and one subscriber socket.  All publishers and all subscribers are connected to the same address and port.   This means that I effectively have a reliable broadcast channel (epgm ensures that all receivers get the message, in the correct order, etc.).  I wrote a simple control layer/protocol on top that allows all nodes to communicate some minimal control information, which in your case might include zmq URIs.  That should allow router-router connections.
>
> I don't know if the IANA has any addresses in the ad-hoc block (http://en.wikipedia.org/wiki/Multicast_address#AD-HOC_block) which are open to individual developers, but that is another thing to look at; if you can get permission to use an address in there, then the pub-sub trick would work across the internet.
>
> Thanks,
> Cem Karan
>
> On Jul 28, 2012, at 8:11 AM, Bjorn Reese wrote:
>
>> On 2012-07-28 13:02, Cem Karan wrote:
>>> Can you describe your system a bit more?  Are the peers on the same LAN, or do they need to peer across the internet?  Do the peers already know about each other (as in, a small group with well-known identities), or do they have to discover each other?  etc...
>>
>> Initially, all peers are on the same LAN where they discover each other
>> via ZeroConf (Bonjour). These local peers constitutes one cluster.
>> Later this has to be extended to several clusters communicating over
>> WAN. The exact means for this has not been determined yet.
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From arkrost at gmail.com  Mon Jul 30 12:22:30 2012
From: arkrost at gmail.com (Arkady Rost)
Date: Mon, 30 Jul 2012 14:22:30 +0400
Subject: [zeromq-dev] msvc and pgm
Message-ID: <CAP8nFi6ZqKLo-uDFO0kaL+xpTE0P4X4+30jGPtWb2FkGL6f3iw@mail.gmail.com>

Hi! I have problems with building zeromq 3.2 with enabled openpgm support
under windows 7 in msvc 2010. Here is compiling log:
http://pastebin.com/QBTvyZDv.  I've just open WithOpenPGM configuration and
tried to build a solution. Am I missing some steps?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/ca75149c/attachment.htm 

From Kevin.Doherty at mlp.com  Mon Jul 30 12:43:59 2012
From: Kevin.Doherty at mlp.com (Doherty, Kevin)
Date: Mon, 30 Jul 2012 06:43:59 -0400
Subject: [zeromq-dev] poller lifecycle
Message-ID: <49D0E89A1F0DC445A55F4938115FC3F039535BCA22@EXCHUS100.AD.MLP.COM>

Folks,
             commenting out the line below results in the exception shown.
When creating a new socket after closing the old.
Unregistering the closed socket from the poller object and adding the new one to the same poller object does not seem to work.
The poll/send complains about an invaid argument ?
(Creating a new poller object is fine.)

                                    params.items.unregister(params.zmqPubSocket);
                                    params.zmqPubSocket.close();
                                    params.zmqPubSocket = ((ZMQ.Context)zmqContext).socket (ZMQ.PUSH);
                                    //**************** params.items = ((ZMQ.Context)zmqContext).poller();
                                    params.zmqPubSocket.setLinger(0);
                                    params.items.register(params.zmqPubSocket, ZMQ.Poller.POLLOUT);
                                    params.items.register(params.zmqPubSocket, ZMQ.Poller.POLLERR);
                                    params.zmqPubSocket.bind (params.sendTo);
                                    try{Thread.sleep(5000);}catch(Exception e){}
                                    params.items.poll(1000000);  <<<<<<<<<<<<<<<<<<<<<<<<<<<


zmq PUSH send failed 22
Connect: zmq throwing
org.zeromq.ZMQException: Invalid argument(0x16)
        at org.zeromq.ZMQ$Poller.run_poll(Native Method)
        at org.zeromq.ZMQ$Poller.poll(ZMQ.java:1298)


######################################################################
The information contained in this communication is confidential and
may contain information that is privileged or exempt from disclosure
under applicable law. If you are not a named addressee, please notify
the sender immediately and delete this email from your system.
If you have received this communication, and are not a named
recipient, you are hereby notified that any dissemination,
distribution or copying of this communication is strictly prohibited.
######################################################################
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/f57cbdfe/attachment.htm 

From cfkaran2 at gmail.com  Mon Jul 30 12:52:15 2012
From: cfkaran2 at gmail.com (Cem Karan)
Date: Mon, 30 Jul 2012 06:52:15 -0400
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <501646EB.9070807@mail1.stofanet.dk>
References: <5013C451.1080000@mail1.stofanet.dk>
	<91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
	<5013D703.60502@mail1.stofanet.dk>
	<D9AD4F80-1DA3-4E96-9002-05F184FACC00@gmail.com>
	<501646EB.9070807@mail1.stofanet.dk>
Message-ID: <5BCF2D34-9635-4E32-BE26-5122E86612E5@gmail.com>

I'm afraid that pub-sub was the only way I could find to fix my problem, and that building some control structures on top of it worked for me.  On the other hand, my use case is probably very different than yours.

The only other suggestion I can make would be to define some kind of control on top of your router-router connections and basically ping the remote end until you get a message back.  If you put that in a moderately tight loop, say, one ping message every 10 milliseconds or so, you'll eventually get through, and the fact that you get a return tells you the connection is up.  I know that that is not ideal because you'll end up with a long queue of pings that you can't get rid of after you've received a reply, but it is all I can offer as a suggestion.

Good luck,
Cem Karan

On Jul 30, 2012, at 4:33 AM, Bjorn Reese wrote:

> We are straying off-topic.
> 
> I also use publish-subscribe in my system, but I need router-router for
> point-to-point interaction.
> 
> So my original problem about messages being silently discarded before
> the identity exchange is finalized is still relevant.
> 
> Let me elaborate on the original problem with an example. The example
> is against https://github.com/zeromq/zeromq3-x and error handling
> has been removed for brevity (but it is present in my original code)
> 
>   void *client = zmq_socket(context, ZMQ_ROUTER);
>   zmq_setsockopt(client, ZMQ_IDENTITY, identity, strlen(identity));
>   zmq_connect(client, endpoint);
> 
>   sleep(1); /* Wait for identity exchange to finish */
> 
>   zmq_msg_t message;
>   zmq_msg_init_size(&message, strlen(data));
>   memcpy(zmq_msg_data(&message), data, strlen(data));
>   int rc = zmq_sendmsg(client, &message, 0);
>   assert(rc == 0);
> 
> If the sleep() function above is removed, then the subsequent message
> is silently ignored.
> 
> 
> On 2012-07-28 20:48, Cem Karan wrote:
>> OK, you're doing something similar to what I'm trying to do, except that you're planning on expanding to a WAN.  I'm going to outline what I'm doing, but I want to note that what I have are a number of pieces of test code that let me make sure that the way I'm using zmq makes sense.  I haven't tried to push large amounts of data through yet, so it may break.  With that disclaimer in mind...
>> 
>> I started out with router-router connections, but switched to pub-sub over epgm and administratively scoped multicast addresses (http://en.wikipedia.org/wiki/Multicast_address#Administratively_Scoped_IPv4_Multicast_addresses) for precisely the reasons you outlined in your original email; it isn't easy distributing the addresses.  In my code, each node has one publisher socket and one subscriber socket.  All publishers and all subscribers are connected to the same address and port.   This means that I effectively have a reliable broadcast channel (epgm ensures that all receivers get the message, in the correct order, etc.).  I wrote a simple control layer/protocol on top that allows all nodes to communicate some minimal control information, which in your case might include zmq URIs.  That should allow router-router connections.
>> 
>> I don't know if the IANA has any addresses in the ad-hoc block (http://en.wikipedia.org/wiki/Multicast_address#AD-HOC_block) which are open to individual developers, but that is another thing to look at; if you can get permission to use an address in there, then the pub-sub trick would work across the internet.
>> 
>> Thanks,
>> Cem Karan
>> 
>> On Jul 28, 2012, at 8:11 AM, Bjorn Reese wrote:
>> 
>>> On 2012-07-28 13:02, Cem Karan wrote:
>>>> Can you describe your system a bit more?  Are the peers on the same LAN, or do they need to peer across the internet?  Do the peers already know about each other (as in, a small group with well-known identities), or do they have to discover each other?  etc...
>>> 
>>> Initially, all peers are on the same LAN where they discover each other
>>> via ZeroConf (Bonjour). These local peers constitutes one cluster.
>>> Later this has to be extended to several clusters communicating over
>>> WAN. The exact means for this has not been determined yet.
>>> 
>>> 
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1581 bytes
Desc: not available
Url : http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/5ac8523b/attachment.bin 

From breese at mail1.stofanet.dk  Mon Jul 30 14:26:43 2012
From: breese at mail1.stofanet.dk (Bjorn Reese)
Date: Mon, 30 Jul 2012 14:26:43 +0200
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <5BCF2D34-9635-4E32-BE26-5122E86612E5@gmail.com>
References: <5013C451.1080000@mail1.stofanet.dk>
	<91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
	<5013D703.60502@mail1.stofanet.dk>
	<D9AD4F80-1DA3-4E96-9002-05F184FACC00@gmail.com>
	<501646EB.9070807@mail1.stofanet.dk>
	<5BCF2D34-9635-4E32-BE26-5122E86612E5@gmail.com>
Message-ID: <50167D83.5010007@mail1.stofanet.dk>

On 2012-07-30 12:52, Cem Karan wrote:
> The only other suggestion I can make would be to define some kind of control on top of your router-router connections and basically ping the remote end until you get a message back.  If you put that in a moderately tight loop, say, one ping message every 10 milliseconds or so, you'll eventually get through, and the fact that you get a return tells you the connection is up.  I know that that is not ideal because you'll end up with a long queue of pings that you can't get rid of after you've received a reply, but it is all I can offer as a suggestion.

I do not think this is necessary, although I appreciate the suggestion.

ZMQ knows when it has received the identity of the peer. It is just a 
question of letting it act upon this event (in one of the three ways
that I have suggested in my original post.)

If the ZMQ developers can agree on what the behavior should be, then I
can implement it.

From steven.mccoy at miru.hk  Mon Jul 30 15:33:03 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Mon, 30 Jul 2012 09:33:03 -0400
Subject: [zeromq-dev] msvc and pgm
In-Reply-To: <CAP8nFi6ZqKLo-uDFO0kaL+xpTE0P4X4+30jGPtWb2FkGL6f3iw@mail.gmail.com>
References: <CAP8nFi6ZqKLo-uDFO0kaL+xpTE0P4X4+30jGPtWb2FkGL6f3iw@mail.gmail.com>
Message-ID: <CAGFWLihwGF5GYKn3LnZJMW1=c34P250Oab7sGaefjDms1y8BrQ@mail.gmail.com>

On 30 July 2012 06:22, Arkady Rost <arkrost at gmail.com> wrote:

> Hi! I have problems with building zeromq 3.2 with enabled openpgm support
> under windows 7 in msvc 2010. Here is compiling log:
> http://pastebin.com/QBTvyZDv.  I've just open WithOpenPGM configuration
> and tried to build a solution. Am I missing some steps?
>

That build target is for the build bot, you will have to modify the path
configuration for it to work in your environment.

I use *CMake *to make the packages available here:

http://www.zeromq.org/distro:microsoft-windows

The CMakeLists.txt is included in some of the repo's but still requires a
few minor steps before working out of the box, but pretty straightforward.

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/bca1e9ad/attachment.htm 

From arkrost at gmail.com  Mon Jul 30 17:34:09 2012
From: arkrost at gmail.com (Arkady Rost)
Date: Mon, 30 Jul 2012 19:34:09 +0400
Subject: [zeromq-dev] msvc and pgm
In-Reply-To: <CAGFWLihwGF5GYKn3LnZJMW1=c34P250Oab7sGaefjDms1y8BrQ@mail.gmail.com>
References: <CAP8nFi6ZqKLo-uDFO0kaL+xpTE0P4X4+30jGPtWb2FkGL6f3iw@mail.gmail.com>
	<CAGFWLihwGF5GYKn3LnZJMW1=c34P250Oab7sGaefjDms1y8BrQ@mail.gmail.com>
Message-ID: <CAP8nFi5F-6JC0WbtRGGJFQadBn-v+VBG0YoTWzM2iB4_=0pyNA@mail.gmail.com>

Thanks for reply. I've managed to solve this problem. I've download win64
openpgm installer from  http://miru.hk/openpgm/. During compilation zeromq
with "WithOpenPGM" configuration added  ${OpenPGM_install_dir}/include to
IncludeDirectories and   ${OpenPGM_install_dir}/lib to LibraryDirectories.
Then I recompiled jzmq and got working bindings.

I don't understand how to use
"zeromq-3.2.0\foreign\openpgm\libpgm-5.1.118~dfsg.tar.gz" that is mentioned
on this page  http://www.zeromq.org/area:download/. I think it would be
great if someone could update installation instructions.

Yours,
  Arkady Rost
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/4d8e346b/attachment.htm 

From iggy.philly at hotmail.com  Mon Jul 30 17:38:39 2012
From: iggy.philly at hotmail.com (Iggy Philly)
Date: Mon, 30 Jul 2012 11:38:39 -0400
Subject: [zeromq-dev] TCP vs IPC on common host
Message-ID: <SNT140-W5960C8D611975E7A7AFC3EF9C60@phx.gbl>





I'm trying to understand the differences between using TCP and IPC between processes on the same linux host. I've seen various indications that the TCP and IPC infrastructures are the same so I'm wondering if there is any significant performance difference between the two protocols on the same host machine. Does the TCP protocol still go through the network stack if both ends are on the same host? Or could I expect performance to be similar? Thanks! 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/ac5e26d4/attachment.htm 

From lists at chuckremes.com  Mon Jul 30 17:51:13 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 30 Jul 2012 10:51:13 -0500
Subject: [zeromq-dev] TCP vs IPC on common host
In-Reply-To: <SNT140-W5960C8D611975E7A7AFC3EF9C60@phx.gbl>
References: <SNT140-W5960C8D611975E7A7AFC3EF9C60@phx.gbl>
Message-ID: <E3F5456D-D9DA-4143-A0BC-C941CA94A70F@chuckremes.com>

On Jul 30, 2012, at 10:38 AM, Iggy Philly wrote:

> 
> I'm trying to understand the differences between using TCP and IPC between processes on the same linux host. I've seen various indications that the TCP and IPC infrastructures are the same so I'm wondering if there is any significant performance difference between the two protocols on the same host machine. Does the TCP protocol still go through the network stack if both ends are on the same host? Or could I expect performance to be similar?

This is a great example of a question that could be answered by running the "perf" examples and trying out the different transports to see any differences.

That's the only way to get a good answer for your environment. 

On Windows, IPC is *currently* handled by opening up a local TCP connection through loopback. However, that could change in a future release and result in a performance change. 

Same is true for Linux; more recent releases have a highly optimized kernel code path for IPC than for TCP to the same host. Try it and see. 

On OSX, IPC is handled via UNIX domain sockets so that the performance relative to TCP is pretty similar. Future releases of OSX may change this.

You will also see performance differences depending on the size of your messages and the transport choice. Try it and see for *your app.*

When doing any benchmarking, it's important that your test run for 30 or more seconds. Any "performance test" that completes in a few seconds will be dominated by startup & shutdown time and therefore give you bogus numbers, so make sure it runs longer.

cr


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/e1fc4010/attachment.htm 

From steven.mccoy at miru.hk  Mon Jul 30 18:46:56 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Mon, 30 Jul 2012 12:46:56 -0400
Subject: [zeromq-dev] msvc and pgm
In-Reply-To: <CAP8nFi5F-6JC0WbtRGGJFQadBn-v+VBG0YoTWzM2iB4_=0pyNA@mail.gmail.com>
References: <CAP8nFi6ZqKLo-uDFO0kaL+xpTE0P4X4+30jGPtWb2FkGL6f3iw@mail.gmail.com>
	<CAGFWLihwGF5GYKn3LnZJMW1=c34P250Oab7sGaefjDms1y8BrQ@mail.gmail.com>
	<CAP8nFi5F-6JC0WbtRGGJFQadBn-v+VBG0YoTWzM2iB4_=0pyNA@mail.gmail.com>
Message-ID: <CAGFWLig7ofc+K1fvKUp_iQPvdqCah3MYrYC6zKebCCZzC+F9Jg@mail.gmail.com>

On 30 July 2012 11:34, Arkady Rost <arkrost at gmail.com> wrote:

> Thanks for reply. I've managed to solve this problem. I've download win64
> openpgm installer from  http://miru.hk/openpgm/. During compilation
> zeromq with "WithOpenPGM" configuration added  ${OpenPGM_install_dir}/include
> to IncludeDirectories and   ${OpenPGM_install_dir}/lib to
> LibraryDirectories. Then I recompiled jzmq and got working bindings.
>
>
There's also ZeroMQ with OpenPGM and JZMQ installers available, the Java
bindings are hiding a bit:

http://miru.hk/zeromq/


> I don't understand how to use
> "zeromq-3.2.0\foreign\openpgm\libpgm-5.1.118~dfsg.tar.gz" that is mentioned
> on this page  http://www.zeromq.org/area:download/. I think it would be
> great if someone could update installation instructions.
>
>
Those instructions are generic for every platform but Windows which
requires additional work for non-trivial builds.

Building OpenPGM is outside the scope of ZeroMQ but documented on the
OpenPGM site:

https://code.google.com/p/openpgm/wiki/OpenPgm5CReferenceBuildLibraryWindows

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/ab16a274/attachment.htm 

From pelletier.michel at gmail.com  Mon Jul 30 19:00:48 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 30 Jul 2012 10:00:48 -0700
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <50167D83.5010007@mail1.stofanet.dk>
References: <5013C451.1080000@mail1.stofanet.dk>
	<91C79A3D-F595-4567-B006-86F71B240B93@gmail.com>
	<5013D703.60502@mail1.stofanet.dk>
	<D9AD4F80-1DA3-4E96-9002-05F184FACC00@gmail.com>
	<501646EB.9070807@mail1.stofanet.dk>
	<5BCF2D34-9635-4E32-BE26-5122E86612E5@gmail.com>
	<50167D83.5010007@mail1.stofanet.dk>
Message-ID: <CACxu=v+egQp8=_U+KexSfar1Hoj-Gop_5sKyDXOMeJuJq5e+yQ@mail.gmail.com>

Initially I think it makes sense to implement this as a monitor event
so that the signaling is out of band.

As for having the 0mq developers "agree", don't worry about that.
Follow the process here and your feature will get merged:

http://www.zeromq.org/docs:contributing

-Michel

On Mon, Jul 30, 2012 at 5:26 AM, Bjorn Reese <breese at mail1.stofanet.dk> wrote:
> On 2012-07-30 12:52, Cem Karan wrote:
>> The only other suggestion I can make would be to define some kind of control on top of your router-router connections and basically ping the remote end until you get a message back.  If you put that in a moderately tight loop, say, one ping message every 10 milliseconds or so, you'll eventually get through, and the fact that you get a return tells you the connection is up.  I know that that is not ideal because you'll end up with a long queue of pings that you can't get rid of after you've received a reply, but it is all I can offer as a suggestion.
>
> I do not think this is necessary, although I appreciate the suggestion.
>
> ZMQ knows when it has received the identity of the peer. It is just a
> question of letting it act upon this event (in one of the three ways
> that I have suggested in my original post.)
>
> If the ZMQ developers can agree on what the behavior should be, then I
> can implement it.
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From diffuser78 at gmail.com  Mon Jul 30 20:48:25 2012
From: diffuser78 at gmail.com (diffuser78)
Date: Mon, 30 Jul 2012 11:48:25 -0700
Subject: [zeromq-dev] looking inside ZMQ queue
In-Reply-To: <CAJT18ia4C+2HVOECzwJCmcyRu6huEFB0GvcqEPx=Z352iu5c6A@mail.gmail.com>
References: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
	<CAEYnEwHYZoNFciiiU0SetR7HFR0ctaaJyEwWo7dGuVcAxndUnQ@mail.gmail.com>
	<CAKni0ef2aWzz9zzJ_-qSnV5cDMYcnX=OqLahSPhe0wk2YmQsNg@mail.gmail.com>
	<CAJT18ia4C+2HVOECzwJCmcyRu6huEFB0GvcqEPx=Z352iu5c6A@mail.gmail.com>
Message-ID: <CAKni0efsKKYbDZB6_VP5oRP4+Qv6G2tQRMVJD2EKwAmr8UmmMQ@mail.gmail.com>

Ok, so once ZMQ sends to transport, zmq removes it from its internal queue.
Is there a way to know when would ZMQ do that ? Or may be build one ? The
customers of my lib want TCP like semantics and I need to tell if the
buffer is copied successfully to the kernel.

Any help on this is greatly appreciated.

Thanks.

DJ
On Fri, Jul 27, 2012 at 3:46 PM, Bennie Kloosteman <bklooste at gmail.com>wrote:

> Like most systems ZeroMQ will just send the message to the transport when
> the transport has received the message ( eg TCP/IP )it is removed from
> ZeroMQ. There are no guarantees it has even been sent , yet alone that the
> transport layer on the other side has the message.  This is pretty standard
> TCP/IP provides no guarantees the other service has received the message (
> only the receiver TCp/IP stack) so you need to code this in your
> application logic.
>
> Also if you are persisting you do not need a light weight system ,
> persistence is always slow and heavy , killing any performance ( Normally
> non persistent messaging systems do at least 10* the messages / second to
> persistent ones) . For persistence I would use a queue like MSMQ or the IBM
> equivalent   . Queues also help in knowing if the other side has received
> it.
>
> Ben
>
>
>
> On Sat, Jul 28, 2012 at 2:27 AM, diffuser78 <diffuser78 at gmail.com> wrote:
>
>> Sorry about my ignorance and laziness. I read this on the FAQ page:
>>
>>
>> *How do I determine how many messages are in queue?*
>>
>> *This isn't possible. At any given time a message may be in the ?MQ
>> sender queue, the sender's kernel buffer, on the wire, in the receiver's
>> kernel buffer or in the receiver's ?MQ receiver queue. Furthermore, a ?MQ
>> socket can bind and/or connect to many peers. Each peer may have different
>> performance characteristics and therefore a different queue depth. Any
>> "queue depth" number is almost certainly wrong, so rather than provide
>> incorrect information the library avoids providing any view into this data.
>> *
>> I had a related question but not same as above.
>>
>> When ZMQ sent this message over TCP to the receivier and when receiver
>> did a FIN/ACK indicating that it is ACK'ing to the finish of the
>> tranmission of byte transfer, it indicates that the transport layer on the
>> other end got the message. Does ZMQ keep tab of this ? When does sender ZMQ
>> removes this message from its internal buffer. *I am only interested in
>> knowing (in YES or NO) whether this message is in sender zmq buffer or not
>> *. Is that possible ? If you can point me out, I can create a patch for
>> it and send for review.
>>
>> Any thoughts are greatly appreciated.
>>
>> DJ
>>
>> On Fri, Jul 27, 2012 at 11:16 AM, Ian Barber <ian.barber at gmail.com>wrote:
>>
>>> On Fri, Jul 27, 2012 at 6:26 PM, diffuser78 <diffuser78 at gmail.com>
>>> wrote:
>>> > What I want to know is that, ZMQ might have returned success upon
>>> buffering
>>> > the message. But what if message is still in the queue and has not been
>>> > delivered to the receiver for some reasons (say slow receiver). Is
>>> there a
>>> > way I can peek into ZMQ queue to find out if my message M that I had
>>> sent in
>>> > a non blocking way was indeed sent on the wire successfully? My lib
>>> needs to
>>> > acknowledge to the app if there was an error after some time T for it
>>> to
>>> > timeout on the message M.
>>>
>>> It could be lost in the network, or the other side could crash while
>>> processing. The best way is to send an acknowledgement back from the
>>> receiver (also using ZeroMQ) and use that to determine whether the
>>> other side has received and processed it.
>>>
>>> Ian
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>>
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>
>>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/73e505cc/attachment.htm 

From lists at chuckremes.com  Mon Jul 30 21:01:53 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 30 Jul 2012 14:01:53 -0500
Subject: [zeromq-dev] looking inside ZMQ queue
In-Reply-To: <CAKni0efsKKYbDZB6_VP5oRP4+Qv6G2tQRMVJD2EKwAmr8UmmMQ@mail.gmail.com>
References: <CAKni0edkS1npvy6XTE7c6eLvU2aT602SNQrguKPmCs9Sr+NE_Q@mail.gmail.com>
	<CAEYnEwHYZoNFciiiU0SetR7HFR0ctaaJyEwWo7dGuVcAxndUnQ@mail.gmail.com>
	<CAKni0ef2aWzz9zzJ_-qSnV5cDMYcnX=OqLahSPhe0wk2YmQsNg@mail.gmail.com>
	<CAJT18ia4C+2HVOECzwJCmcyRu6huEFB0GvcqEPx=Z352iu5c6A@mail.gmail.com>
	<CAKni0efsKKYbDZB6_VP5oRP4+Qv6G2tQRMVJD2EKwAmr8UmmMQ@mail.gmail.com>
Message-ID: <D003D1E9-F94B-403E-84AE-B8A3B44CD1B2@chuckremes.com>


On Jul 30, 2012, at 1:48 PM, diffuser78 wrote:

> Ok, so once ZMQ sends to transport, zmq removes it from its internal queue. Is there a way to know when would ZMQ do that ? Or may be build one ? The customers of my lib want TCP like semantics and I need to tell if the buffer is copied successfully to the kernel.
>  
> Any help on this is greatly appreciated.
>  
> Thanks.

If your customers need TCP-like semantics, then why not create FIN/ACK (and others) as zeromq messages in your application?

Every person who has responded to this thread has already told you how to solve this. The receiver must ACK the message. Then, and *only* then, will you know that the message has been sent and received.

cr



From diffuser78 at gmail.com  Mon Jul 30 21:05:48 2012
From: diffuser78 at gmail.com (diffuser78)
Date: Mon, 30 Jul 2012 12:05:48 -0700
Subject: [zeromq-dev] How to get reliability in pub-sub pattern
Message-ID: <CAKni0edE9ONsvLcxFawHmMTVKW2xcQQ03g3Gbf66RnKQao98AA@mail.gmail.com>

Hi,


From lists at chuckremes.com  Mon Jul 30 21:19:45 2012
From: lists at chuckremes.com (Chuck Remes)
Date: Mon, 30 Jul 2012 14:19:45 -0500
Subject: [zeromq-dev] How to get reliability in pub-sub pattern
In-Reply-To: <CAKni0edE9ONsvLcxFawHmMTVKW2xcQQ03g3Gbf66RnKQao98AA@mail.gmail.com>
References: <CAKni0edE9ONsvLcxFawHmMTVKW2xcQQ03g3Gbf66RnKQao98AA@mail.gmail.com>
Message-ID: <DBE9FC90-2464-4EF0-BC4E-7891924A4657@chuckremes.com>

On Jul 30, 2012, at 2:05 PM, diffuser78 wrote:

> Hi,
>  
> From ZMQ guide I see that pub-sub pattern does not offer reliability, i.e. publisher doesn't know if the subscriber has been consuming messages.
>  
> What other pattern would you recommend in pub-sub scenario where a publisher must know the details of subscribers and how many messages subscribers have consumed ?

I should write a wiki article about an idea/opinion that I have. I brought it up at the Chicago meetup from a few months ago.

The only socket types that we should encourage people to use are DEALER/ROUTER (and there was some discussion that this could be reduced to just ROUTER sockets). With DEALER/ROUTER, it is (hopefully) obvious how to do reliable communications between any number of peers. The patterns in the guide illustrate this nicely and in most (all?) cases use the DEALER/ROUTER socket types.

For pub/sub, if you need reliability then you absolutely need bidirectional communications at your application level. The PUB/SUB sockets in the library do not offer this, so now you are back to DEALER/ROUTER.

You could mimic reliable pub/sub by using two sets of sockets where the subscribers on the first set would publish the sequence numbers that have been received (or a range). Much like a sliding-window protocol, you could cheaply "ack" lots of messages. But now you need to build in additional intelligence to all subscribers so that they know to ignore duplicate sequence numbers. Plus, you have the complexity of creating a truly unique sequence number in a distributed system; that generator is oftentimes a bottle neck.

Good luck. Let us know what you figure out.

cr



From pelletier.michel at gmail.com  Mon Jul 30 23:06:47 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 30 Jul 2012 14:06:47 -0700
Subject: [zeromq-dev] How to get reliability in pub-sub pattern
In-Reply-To: <DBE9FC90-2464-4EF0-BC4E-7891924A4657@chuckremes.com>
References: <CAKni0edE9ONsvLcxFawHmMTVKW2xcQQ03g3Gbf66RnKQao98AA@mail.gmail.com>
	<DBE9FC90-2464-4EF0-BC4E-7891924A4657@chuckremes.com>
Message-ID: <CACxu=vJuJr7sTKfiGjkzt6Ybe3bvDByyPE8_BA=ka6SjBDkCfg@mail.gmail.com>

Also check out:

http://zguide.zeromq.org/page:all#A-Shared-Key-Value-Cache-Clone-Pattern

-Michel

On Mon, Jul 30, 2012 at 12:19 PM, Chuck Remes <lists at chuckremes.com> wrote:
> On Jul 30, 2012, at 2:05 PM, diffuser78 wrote:
>
>> Hi,
>>
>> From ZMQ guide I see that pub-sub pattern does not offer reliability, i.e. publisher doesn't know if the subscriber has been consuming messages.
>>
>> What other pattern would you recommend in pub-sub scenario where a publisher must know the details of subscribers and how many messages subscribers have consumed ?
>
> I should write a wiki article about an idea/opinion that I have. I brought it up at the Chicago meetup from a few months ago.
>
> The only socket types that we should encourage people to use are DEALER/ROUTER (and there was some discussion that this could be reduced to just ROUTER sockets). With DEALER/ROUTER, it is (hopefully) obvious how to do reliable communications between any number of peers. The patterns in the guide illustrate this nicely and in most (all?) cases use the DEALER/ROUTER socket types.
>
> For pub/sub, if you need reliability then you absolutely need bidirectional communications at your application level. The PUB/SUB sockets in the library do not offer this, so now you are back to DEALER/ROUTER.
>
> You could mimic reliable pub/sub by using two sets of sockets where the subscribers on the first set would publish the sequence numbers that have been received (or a range). Much like a sliding-window protocol, you could cheaply "ack" lots of messages. But now you need to build in additional intelligence to all subscribers so that they know to ignore duplicate sequence numbers. Plus, you have the complexity of creating a truly unique sequence number in a distributed system; that generator is oftentimes a bottle neck.
>
> Good luck. Let us know what you figure out.
>
> cr
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From pelletier.michel at gmail.com  Mon Jul 30 23:26:31 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 30 Jul 2012 14:26:31 -0700
Subject: [zeromq-dev] TCP vs IPC on common host
In-Reply-To: <E3F5456D-D9DA-4143-A0BC-C941CA94A70F@chuckremes.com>
References: <SNT140-W5960C8D611975E7A7AFC3EF9C60@phx.gbl>
	<E3F5456D-D9DA-4143-A0BC-C941CA94A70F@chuckremes.com>
Message-ID: <CACxu=vK0frC56zYxy7mdQnu26p148W87SEkXcsQkuXXCCYFhYg@mail.gmail.com>

On Mon, Jul 30, 2012 at 8:51 AM, Chuck Remes <lists at chuckremes.com> wrote:
> On Jul 30, 2012, at 10:38 AM, Iggy Philly wrote:

> On Windows, IPC is *currently* handled by opening up a local TCP connection
> through loopback. However, that could change in a future release and result
> in a performance change.
>
> Same is true for Linux; more recent releases have a highly optimized kernel
> code path for IPC than for TCP to the same host. Try it and see.

I always thought (and it looks to me by brief code inspection) that
Linux uses domain sockets as well.

>
> On OSX, IPC is handled via UNIX domain sockets so that the performance
> relative to TCP is pretty similar. Future releases of OSX may change this.

-Michel

From steven.mccoy at miru.hk  Tue Jul 31 01:20:54 2012
From: steven.mccoy at miru.hk (Steven McCoy)
Date: Mon, 30 Jul 2012 19:20:54 -0400
Subject: [zeromq-dev] TCP vs IPC on common host
In-Reply-To: <CACxu=vK0frC56zYxy7mdQnu26p148W87SEkXcsQkuXXCCYFhYg@mail.gmail.com>
References: <SNT140-W5960C8D611975E7A7AFC3EF9C60@phx.gbl>
	<E3F5456D-D9DA-4143-A0BC-C941CA94A70F@chuckremes.com>
	<CACxu=vK0frC56zYxy7mdQnu26p148W87SEkXcsQkuXXCCYFhYg@mail.gmail.com>
Message-ID: <CAGFWLihGK4wMsPdKauw1R8JCxVFPJpLAU5NDoQ+h_5AKMfRpvw@mail.gmail.com>

On 30 July 2012 17:26, Michel Pelletier <pelletier.michel at gmail.com> wrote:

> On Mon, Jul 30, 2012 at 8:51 AM, Chuck Remes <lists at chuckremes.com> wrote:
> > On Jul 30, 2012, at 10:38 AM, Iggy Philly wrote:
>
> > On Windows, IPC is *currently* handled by opening up a local TCP
> connection
> > through loopback. However, that could change in a future release and
> result
> > in a performance change.
> >
> > Same is true for Linux; more recent releases have a highly optimized
> kernel
> > code path for IPC than for TCP to the same host. Try it and see.
>
> I always thought (and it looks to me by brief code inspection) that
> Linux uses domain sockets as well.
>
>
I think the point is Unix sockets are faster than TCP because of less
layers to traverse.  Solaris is still I think the only OS that can bump
localhost TCP connections to Unix sockets.

-- 
Steve-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120730/32dbf5d9/attachment.htm 

From stu at compuserve.com  Tue Jul 31 03:01:55 2012
From: stu at compuserve.com (Stuart Brandt)
Date: Mon, 30 Jul 2012 21:01:55 -0400
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
In-Reply-To: <5012EAB7.80500@compuserve.com>
References: <5012EAB7.80500@compuserve.com>
Message-ID: <50172E83.5040006@compuserve.com>

Just a followup that a similar issue appears to exist with ZMQ_ROUTER 
connecting the ZMQ_DEALER. In this case I'm seeing...

Assertion failed: (prefetched_msg.flags () & msg_t::identity) == 0 
(router.cpp:274)
Aborted

I repro'd this one with two Queue devices connecting back-to-back -- the 
2nd Queue's front-end Router connecting to the back-end Dealer of the 
first Queue, then killing and restarting the process with the first 
Queue. Is there some sockopt I'm missing that could be contributing to this?

On 7/27/2012 3:23 PM, Stuart Brandt wrote:
> I get an assertion during a call to zmq_recv on 2 different OSes (OSX
> 10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and
> a recent build off latest at github.com/zeromq/libzmq).
>
> The assertion is:
> Assertion failed: (msg_->flags ()&  msg_t::identity) == 0 (router.cpp:220)
> Aborted
>
> The setup involves:
> 1) the simple message queue broker found at
> https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
> I modified the #include to look for zmq.h rather than zhelpers.h to get
> it to compile with ZMQ 3.x
>
> 2) a minimal server implementation that connects to the broker and does
> a zmq_recv. Code is:
> //
> //  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
> //  Connects to tcp://localhost:5560
> //
> #include<stdio.h>
> #include<zmq.h>
>
> int main (void)
> {
>       char buf[80];   // buffer for zmq_recv call
>       int len;        // len returned by zmq_recv
>       void *ctx;      // ZMQ context
>       void *sock;     // ZMQ socket
>
>       // init ZMQ context, create a REP socket, and connect to ROUTER
>       ctx = zmq_ctx_new();
>       sock = zmq_socket(ctx,ZMQ_REP);
>       zmq_connect (sock, "tcp://localhost:5560");
>
>       // do a basic zmq_recv call...warn if it fails
>       len = zmq_recv(sock,buf,sizeof(buf),0);
>       if (len<  0) {
>           printf("zmq_recv failed - %s\n",zmq_strerror(errno));
>       }
>
>       // Done. Clean up and end
>       zmq_close (sock);
>       zmq_term (ctx);
>       return 0;
> }
>
>
> Steps to repro are start server, start broker, kill broker (spec.
> ctrl-C), restart broker. Server aborts with the above assertion.
>
> Is there something my code should be doing differently?
>
> Thanks....Stuart
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From pelletier.michel at gmail.com  Tue Jul 31 04:39:10 2012
From: pelletier.michel at gmail.com (Michel Pelletier)
Date: Mon, 30 Jul 2012 19:39:10 -0700
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
In-Reply-To: <50172E83.5040006@compuserve.com>
References: <5012EAB7.80500@compuserve.com> <50172E83.5040006@compuserve.com>
Message-ID: <CACxu=vLCF9yC=3EnEq5_EoOLyrJjeGBGGiy5N0sDMr2=vapsUQ@mail.gmail.com>

Just a guess from looking at the code, but it seems like reconnection
and routers may not be playing along so well as the new connection
sends the identity as the first message, and the router does not think
it's in a state where it is expecting that.  It seems like router
should expect an identity at any time due to a reconnection, and reset
it's state accordingly.  If true, I'd def call that a bug.

-Michel

On Mon, Jul 30, 2012 at 6:01 PM, Stuart Brandt <stu at compuserve.com> wrote:
> Just a followup that a similar issue appears to exist with ZMQ_ROUTER
> connecting the ZMQ_DEALER. In this case I'm seeing...
>
> Assertion failed: (prefetched_msg.flags () & msg_t::identity) == 0
> (router.cpp:274)
> Aborted
>
> I repro'd this one with two Queue devices connecting back-to-back -- the
> 2nd Queue's front-end Router connecting to the back-end Dealer of the
> first Queue, then killing and restarting the process with the first
> Queue. Is there some sockopt I'm missing that could be contributing to this?
>
> On 7/27/2012 3:23 PM, Stuart Brandt wrote:
>> I get an assertion during a call to zmq_recv on 2 different OSes (OSX
>> 10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and
>> a recent build off latest at github.com/zeromq/libzmq).
>>
>> The assertion is:
>> Assertion failed: (msg_->flags ()&  msg_t::identity) == 0 (router.cpp:220)
>> Aborted
>>
>> The setup involves:
>> 1) the simple message queue broker found at
>> https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
>> I modified the #include to look for zmq.h rather than zhelpers.h to get
>> it to compile with ZMQ 3.x
>>
>> 2) a minimal server implementation that connects to the broker and does
>> a zmq_recv. Code is:
>> //
>> //  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
>> //  Connects to tcp://localhost:5560
>> //
>> #include<stdio.h>
>> #include<zmq.h>
>>
>> int main (void)
>> {
>>       char buf[80];   // buffer for zmq_recv call
>>       int len;        // len returned by zmq_recv
>>       void *ctx;      // ZMQ context
>>       void *sock;     // ZMQ socket
>>
>>       // init ZMQ context, create a REP socket, and connect to ROUTER
>>       ctx = zmq_ctx_new();
>>       sock = zmq_socket(ctx,ZMQ_REP);
>>       zmq_connect (sock, "tcp://localhost:5560");
>>
>>       // do a basic zmq_recv call...warn if it fails
>>       len = zmq_recv(sock,buf,sizeof(buf),0);
>>       if (len<  0) {
>>           printf("zmq_recv failed - %s\n",zmq_strerror(errno));
>>       }
>>
>>       // Done. Clean up and end
>>       zmq_close (sock);
>>       zmq_term (ctx);
>>       return 0;
>> }
>>
>>
>> Steps to repro are start server, start broker, kill broker (spec.
>> ctrl-C), restart broker. Server aborts with the above assertion.
>>
>> Is there something my code should be doing differently?
>>
>> Thanks....Stuart
>>
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From sebastian.lauwers at gmail.com  Tue Jul 31 09:00:25 2012
From: sebastian.lauwers at gmail.com (Sebastian Lauwers)
Date: Tue, 31 Jul 2012 09:00:25 +0200
Subject: [zeromq-dev] zeromq 3.2 tests fails
In-Reply-To: <CAESgSZpyoA9gAsjh83oD0BPCyG6EY1kcm8TVXtsFpzd=Ti8mnQ@mail.gmail.com>
References: <CAP8nFi7gCbdP9kTp-d6h4rYg7VYmfX4vR76ZgjeZ=b4aSvzhBA@mail.gmail.com>
	<CAESgSZpyoA9gAsjh83oD0BPCyG6EY1kcm8TVXtsFpzd=Ti8mnQ@mail.gmail.com>
Message-ID: <CAESgSZrwBWPWDVPtJ5oL2abX42+SkmMLaTMG=ZjqQhjcq9zSGQ@mail.gmail.com>

On 30 July 2012 08:44, Sebastian Lauwers <sebastian.lauwers at gmail.com> wrote:
> I'm not familiar enough with ZMQ to understand what's going on, so
> I'll leave the rest to someone more experienced.

Looking at this a bit further, this is what I found.

Please note that I haven't become an expert since yesterday, so I'm
mainly just going at this by intuition. Feel free to correct any
mistakes I may have made or point out any wrong assumptions.

In stream::engine::plug(), a call to session_base_t::get_address(),
which rightfully checks whether its addr property is null before
calling a method on it (session_base.cpp:269):
    if (addr)
        return addr->to_string (addr_);
    return -1;

If addr is correctly initialised, everything is fine and dandy: we
simply return the address in string format.

However, addr isn't always set: In some cases, the session_base_t
object is instantiated with addr = NULL (tcp_listener.cpp:105):
    session_base_t *session = session_base_t::create (io_thread, false, socket,
        options, NULL);

The same line is also usedin ipc_listerner.cpp. However, in
socket_base.cpp (line 492 and 529), the session is created with a
non-NULL parameter:
    address_t *paddr = new (std::nothrow) address_t (protocol, address);
    alloc_assert (paddr);
<snip>
    session_base_t *session = session_base_t::create (io_thread, true, this,
        options, paddr);

Now, passing the paddr to the ::create() call was added in February by
Staffan Gim?ker in commit b9fb48f47b8b6f211625e0897c5f52ba5cfe6593,
which added the ability to propagate resolution errors on connect
rather than exclusively from the IO thread.

If this research is correct, then I guess two things could be done
(again, feel free to add other possibilities if you see them, I'm
shooting in the dark here):
- Find a way to properly create the addr object before creating the
session object, or,
- Not use the session to get the address in string format.

Thoughts?

-S.

From arkrost at gmail.com  Tue Jul 31 09:14:21 2012
From: arkrost at gmail.com (Arkady Rost)
Date: Tue, 31 Jul 2012 11:14:21 +0400
Subject: [zeromq-dev] zeromq 3.2 tests fails
In-Reply-To: <CAESgSZrwBWPWDVPtJ5oL2abX42+SkmMLaTMG=ZjqQhjcq9zSGQ@mail.gmail.com>
References: <CAP8nFi7gCbdP9kTp-d6h4rYg7VYmfX4vR76ZgjeZ=b4aSvzhBA@mail.gmail.com>
	<CAESgSZpyoA9gAsjh83oD0BPCyG6EY1kcm8TVXtsFpzd=Ti8mnQ@mail.gmail.com>
	<CAESgSZrwBWPWDVPtJ5oL2abX42+SkmMLaTMG=ZjqQhjcq9zSGQ@mail.gmail.com>
Message-ID: <CAP8nFi65_Tyo75yTCeuRUkOQ+N1ANgp76tsVf0U1Z8hHK8wv7g@mail.gmail.com>

Extremely apologize for not replying. I'm away from my home pc where I have
problems with tests. In this evening I'll manage to test a bit more. Thank
you for a great response.

--
Arkady Rost
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120731/b2543811/attachment.htm 

From sikamani at ge.com  Tue Jul 31 02:14:33 2012
From: sikamani at ge.com (Sikamani, Kothandaraman (GE Global Research, consultant))
Date: Tue, 31 Jul 2012 00:14:33 +0000
Subject: [zeromq-dev] Cannot build a working JZMQ on Windows: which git
 branch to take?
Message-ID: <52DE3381F9F917438EBEBE0C4530CE62E1B2@CINMBCNA06.e2k.ad.ge.com>

After I build using MS Visiual Studio 64 bit version support with for ZeroMQ Im able to get libzmq.dll under
C:\Code\zeromq-zeromq3\lib

But when I try for Java wrapper to get build jzmq.dll in same Visiual Studio
 I get following erros

1>------ Rebuild All started: Project: jzmq, Configuration: Release Win32 ------
1>Build started 7/30/2012 4:57:52 PM.
1>_PrepareForClean:
1>  Deleting file "Release\jzmq.lastbuildstate".
1>InitializeBuildStatus:
1>  Touching "Release\jzmq.unsuccessfulbuild".
1>PreBuildEvent:
1>  Description: Compiling Java classes
1>          1 file(s) copied.
1>CustomBuild:
1>  Generating JNI header
1>  Generating JNI header
1>  Generating JNI header
1>  Generating JNI header
1>  Generating JNI header
1>  Generating JNI header
1>  Generating JNI header
1>ClCompile:
1>  Context.cpp
1>..\..\..\src\Context.cpp(22): fatal error C1083: Cannot open include file: 'zmq.h': No such file or directory
1>  Poller.cpp
1>..\..\..\src\Poller.cpp(22): fatal error C1083: Cannot open include file: 'zmq.h': No such file or directory
1>  Socket.cpp
1>..\..\..\src\Socket.cpp(23): fatal error C1083: Cannot open include file: 'zmq.h': No such file or directory
1>  util.cpp
1>..\..\..\src\util.cpp(22): fatal error C1083: Cannot open include file: 'zmq.h': No such file or directory
1>  ZMQ.cpp
1>..\..\..\src\ZMQ.cpp(22): fatal error C1083: Cannot open include file: 'zmq.h': No such file or directory
1>
1>Build FAILED.
1>
1>Time Elapsed 00:00:05.06
2>------ Rebuild All started: Project: remote_lat, Configuration: Release Win32 ------
3>------ Rebuild All started: Project: local_lat, Configuration: Release Win32 ------
4>------ Rebuild All started: Project: local_thr, Configuration: Release Win32 ------
5>------ Rebuild All started: Project: remote_thr, Configuration: Release Win32 ------
3>Build started 7/30/2012 4:57:57 PM.
3>_PrepareForClean:
3>  Deleting file "Release\local_lat.lastbuildstate".
2>Build started 7/30/2012 4:57:57 PM.
2>_PrepareForClean:
2>  Deleting file "Release\remote_lat.lastbuildstate".
4>Build started 7/30/2012 4:57:58 PM.
4>_PrepareForClean:
4>  Deleting file "Release\local_thr.lastbuildstate".
2>InitializeBuildStatus:
4>InitializeBuildStatus:
4>  Creating "Release\local_thr.unsuccessfulbuild" because "AlwaysCreate" was specified.
4>PreBuildEvent:
4>  Description: Compiling Java classes
2>  Creating "Release\remote_lat.unsuccessfulbuild" because "AlwaysCreate" was specified.
2>PreBuildEvent:
2>  Description: Compiling Java classes
3>InitializeBuildStatus:
3>  Creating "Release\local_lat.unsuccessfulbuild" because "AlwaysCreate" was specified.
3>PreBuildEvent:
3>  Description: Compiling Java classes
5>Build started 7/30/2012 4:57:58 PM.
5>_PrepareForClean:
5>  Deleting file "Release\remote_thr.lastbuildstate".
5>InitializeBuildStatus:
5>  Creating "Release\remote_thr.unsuccessfulbuild" because "AlwaysCreate" was specified.
5>PreBuildEvent:
5>  Description: Compiling Java classes
4>FinalizeBuildStatus:
4>  Deleting file "Release\local_thr.unsuccessfulbuild".
4>  Touching "Release\local_thr.lastbuildstate".
4>
4>Build succeeded.
4>
4>Time Elapsed 00:00:01.37
3>FinalizeBuildStatus:
3>  Deleting file "Release\local_lat.unsuccessfulbuild".
3>  Touching "Release\local_lat.lastbuildstate".
3>
3>Build succeeded.
3>
3>Time Elapsed 00:00:02.23
2>FinalizeBuildStatus:
2>  Deleting file "Release\remote_lat.unsuccessfulbuild".
2>  Touching "Release\remote_lat.lastbuildstate".
2>
2>Build succeeded.
2>
2>Time Elapsed 00:00:02.06
5>FinalizeBuildStatus:
5>  Deleting file "Release\remote_thr.unsuccessfulbuild".
5>  Touching "Release\remote_thr.lastbuildstate".
5>
5>Build succeeded.
5>
5>Time Elapsed 00:00:01.40
========== Rebuild All: 4 succeeded, 1 failed, 0 skipped ==========


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120731/3f8bbc84/attachment.htm 

From arkrost at gmail.com  Tue Jul 31 10:18:35 2012
From: arkrost at gmail.com (Arkady Rost)
Date: Tue, 31 Jul 2012 12:18:35 +0400
Subject: [zeromq-dev] Cannot build a working JZMQ on Windows: which git
 branch to take?
In-Reply-To: <52DE3381F9F917438EBEBE0C4530CE62E1B2@CINMBCNA06.e2k.ad.ge.com>
References: <52DE3381F9F917438EBEBE0C4530CE62E1B2@CINMBCNA06.e2k.ad.ge.com>
Message-ID: <CAP8nFi6hBfwKR80UT2ZmRWKeRymdd8Cx3+QA5h5MGrBhS8hW+Q@mail.gmail.com>

Hello!
I have same problems yesterday. Please ensure that you've configured your
project correctly. Go to  jzmq Project -> Properties -> VC++ Directories.
Add   ${jdk1.7}/include, ${jdk1.7}/include/win32,
${zeromq_compile_dir}/include to IncludeDirectories and
${zeromq_compile_dir}/lib/x64
to  Library Directories. After that I suppose you'll build jzmq.


--
Arkady Rost
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120731/e4cf2496/attachment.htm 

From hurtonm at gmail.com  Tue Jul 31 12:04:17 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Tue, 31 Jul 2012 12:04:17 +0200
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
In-Reply-To: <5012EAB7.80500@compuserve.com>
References: <5012EAB7.80500@compuserve.com>
Message-ID: <CALdZE=sQJa3NKBc9-V+5zkfyOhA+EVNrm3je3YT646iTvx75SA@mail.gmail.com>

Hi Stuart, could you please file an issue for this?

- Martin

On Fri, Jul 27, 2012 at 9:23 PM, Stuart Brandt <stu at compuserve.com> wrote:
> I get an assertion during a call to zmq_recv on 2 different OSes (OSX
> 10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and
> a recent build off latest at github.com/zeromq/libzmq).
>
> The assertion is:
> Assertion failed: (msg_->flags () & msg_t::identity) == 0 (router.cpp:220)
> Aborted
>
> The setup involves:
> 1) the simple message queue broker found at
> https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
> I modified the #include to look for zmq.h rather than zhelpers.h to get
> it to compile with ZMQ 3.x
>
> 2) a minimal server implementation that connects to the broker and does
> a zmq_recv. Code is:
> //
> //  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
> //  Connects to tcp://localhost:5560
> //
> #include <stdio.h>
> #include <zmq.h>
>
> int main (void)
> {
>      char buf[80];   // buffer for zmq_recv call
>      int len;        // len returned by zmq_recv
>      void *ctx;      // ZMQ context
>      void *sock;     // ZMQ socket
>
>      // init ZMQ context, create a REP socket, and connect to ROUTER
>      ctx = zmq_ctx_new();
>      sock = zmq_socket(ctx,ZMQ_REP);
>      zmq_connect (sock, "tcp://localhost:5560");
>
>      // do a basic zmq_recv call...warn if it fails
>      len = zmq_recv(sock,buf,sizeof(buf),0);
>      if (len < 0) {
>          printf("zmq_recv failed - %s\n",zmq_strerror(errno));
>      }
>
>      // Done. Clean up and end
>      zmq_close (sock);
>      zmq_term (ctx);
>      return 0;
> }
>
>
> Steps to repro are start server, start broker, kill broker (spec.
> ctrl-C), restart broker. Server aborts with the above assertion.
>
> Is there something my code should be doing differently?
>
> Thanks....Stuart
>
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From breese at mail1.stofanet.dk  Tue Jul 31 12:07:35 2012
From: breese at mail1.stofanet.dk (Bjorn Reese)
Date: Tue, 31 Jul 2012 12:07:35 +0200
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <5013C451.1080000@mail1.stofanet.dk>
References: <5013C451.1080000@mail1.stofanet.dk>
Message-ID: <5017AE67.5050808@mail1.stofanet.dk>

On 2012-07-28 12:52, Bjorn Reese wrote:

> 2. Let zmq_send() return EAGAIN until the identity has been exchanged.
>
> 3. Add an event to notify the application that the identity has been
>      exchanged (e.g. via the new zmq_ctx_monitor_event() functionality.)

While delving deeper into this problem, I discovered that the
ZMQ_ROUTER_BEHAVIOR option is more or less what I am looking for.

If this option is enabled, the call to zmq_send() will block until the
identity has been exchanged, thus making the sleep() superfluous. If
zmq_send() is called with ZMQ_DONTWAIT, it will return EAGAIN, which is
exactly as I requested in #2 above.

In the non-blocking case it may still be a good idea to receive an
event when the identity has been exchanged, so I will look further
into that.

From breese at mail1.stofanet.dk  Tue Jul 31 15:21:49 2012
From: breese at mail1.stofanet.dk (Bjorn Reese)
Date: Tue, 31 Jul 2012 15:21:49 +0200
Subject: [zeromq-dev] Router to router delay
In-Reply-To: <5017AE67.5050808@mail1.stofanet.dk>
References: <5013C451.1080000@mail1.stofanet.dk>
	<5017AE67.5050808@mail1.stofanet.dk>
Message-ID: <5017DBED.6040900@mail1.stofanet.dk>

On 2012-07-31 12:07, Bjorn Reese wrote:

> In the non-blocking case it may still be a good idea to receive an
> event when the identity has been exchanged, so I will look further
> into that.

Patch posted in jira (LIBZMQ-407)

From stu at compuserve.com  Tue Jul 31 17:48:14 2012
From: stu at compuserve.com (Stuart Brandt)
Date: Tue, 31 Jul 2012 11:48:14 -0400
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
In-Reply-To: <CALdZE=sQJa3NKBc9-V+5zkfyOhA+EVNrm3je3YT646iTvx75SA@mail.gmail.com>
References: <5012EAB7.80500@compuserve.com>
	<CALdZE=sQJa3NKBc9-V+5zkfyOhA+EVNrm3je3YT646iTvx75SA@mail.gmail.com>
Message-ID: <5017FE3E.9050701@compuserve.com>

Done. LIBZMQ-406

Thanks for confirmation.  Let me know if there's any additional context 
needed.

- Stuart

On 7/31/2012 6:04 AM, Martin Hurton wrote:
> Hi Stuart, could you please file an issue for this?
>
> - Martin
>
> On Fri, Jul 27, 2012 at 9:23 PM, Stuart Brandt <stu at compuserve.com> wrote:
>> I get an assertion during a call to zmq_recv on 2 different OSes (OSX
>> 10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and
>> a recent build off latest at github.com/zeromq/libzmq).
>>
>> The assertion is:
>> Assertion failed: (msg_->flags () & msg_t::identity) == 0 (router.cpp:220)
>> Aborted
>>
>> The setup involves:
>> 1) the simple message queue broker found at
>> https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
>> I modified the #include to look for zmq.h rather than zhelpers.h to get
>> it to compile with ZMQ 3.x
>>
>> 2) a minimal server implementation that connects to the broker and does
>> a zmq_recv. Code is:
>> //
>> //  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
>> //  Connects to tcp://localhost:5560
>> //
>> #include <stdio.h>
>> #include <zmq.h>
>>
>> int main (void)
>> {
>>       char buf[80];   // buffer for zmq_recv call
>>       int len;        // len returned by zmq_recv
>>       void *ctx;      // ZMQ context
>>       void *sock;     // ZMQ socket
>>
>>       // init ZMQ context, create a REP socket, and connect to ROUTER
>>       ctx = zmq_ctx_new();
>>       sock = zmq_socket(ctx,ZMQ_REP);
>>       zmq_connect (sock, "tcp://localhost:5560");
>>
>>       // do a basic zmq_recv call...warn if it fails
>>       len = zmq_recv(sock,buf,sizeof(buf),0);
>>       if (len < 0) {
>>           printf("zmq_recv failed - %s\n",zmq_strerror(errno));
>>       }
>>
>>       // Done. Clean up and end
>>       zmq_close (sock);
>>       zmq_term (ctx);
>>       return 0;
>> }
>>
>>
>> Steps to repro are start server, start broker, kill broker (spec.
>> ctrl-C), restart broker. Server aborts with the above assertion.
>>
>> Is there something my code should be doing differently?
>>
>> Thanks....Stuart
>>
>>
>>
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From ian.barber at gmail.com  Tue Jul 31 18:34:40 2012
From: ian.barber at gmail.com (Ian Barber)
Date: Tue, 31 Jul 2012 17:34:40 +0100
Subject: [zeromq-dev] [PATCH] [cppzmq] Add move constructor to message_t
In-Reply-To: <OF41A37A04.111F3D40-ON80257A4B.002AF330-80257A4B.002B8727@waters.com>
References: <OF21D5BAA1.FE45B344-ON80257A47.00446023-80257A47.0044CCF8@waters.com>
	<CADL5_siQi69HEojq8etz5QrbPtgzG_idDKS9Rba5Nvrd=dHX2Q@mail.gmail.com>
	<CAEYnEwHOQmcaAAYN3TsMVQb+2LTqKwmtJfzVOBEVqtQB3Ep+0g@mail.gmail.com>
	<OF41A37A04.111F3D40-ON80257A4B.002AF330-80257A4B.002B8727@waters.com>
Message-ID: <CAEYnEwH=TNQAAkAHbFGaEVG2oX5wvORF0rigrg3-vKERKCMgeg@mail.gmail.com>

On Mon, Jul 30, 2012 at 8:55 AM, <Richard_Newton at waters.com> wrote:

> Hi Ian,
>
> Thanks, but merged it in where?  I cannot see the changes at
> https://github.com/zeromq/cppzmq.
>
> So I am clear, what is the protocol for submitting patches?  Should I be
> sending the patch to the list for review as well as doing a pull request or
> do you just want the pull request?
>

That's a good question! Sorry, you did the right thing submitting the pull
req, but I hadn't actually merged it - must have just reviewed and made the
merge up in my head.

Thanks,

Ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120731/4da175b1/attachment.htm 

From hurtonm at gmail.com  Tue Jul 31 21:26:36 2012
From: hurtonm at gmail.com (Martin Hurton)
Date: Tue, 31 Jul 2012 21:26:36 +0200
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
In-Reply-To: <5017FE3E.9050701@compuserve.com>
References: <5012EAB7.80500@compuserve.com>
	<CALdZE=sQJa3NKBc9-V+5zkfyOhA+EVNrm3je3YT646iTvx75SA@mail.gmail.com>
	<5017FE3E.9050701@compuserve.com>
Message-ID: <CALdZE=uvMBQjGcv2ZfGyY5dS9rQ17siXnC+Xe8e21RH4=j9WbQ@mail.gmail.com>

Fixed in 9fab9937e516e78e6dbdf890e34f56cf80c9ff14. Can you check and
close the issue if it fixes your problem.

On Tue, Jul 31, 2012 at 5:48 PM, Stuart Brandt <stu at compuserve.com> wrote:
> Done. LIBZMQ-406
>
> Thanks for confirmation.  Let me know if there's any additional context
> needed.
>
> - Stuart
>
> On 7/31/2012 6:04 AM, Martin Hurton wrote:
>> Hi Stuart, could you please file an issue for this?
>>
>> - Martin
>>
>> On Fri, Jul 27, 2012 at 9:23 PM, Stuart Brandt <stu at compuserve.com> wrote:
>>> I get an assertion during a call to zmq_recv on 2 different OSes (OSX
>>> 10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and
>>> a recent build off latest at github.com/zeromq/libzmq).
>>>
>>> The assertion is:
>>> Assertion failed: (msg_->flags () & msg_t::identity) == 0 (router.cpp:220)
>>> Aborted
>>>
>>> The setup involves:
>>> 1) the simple message queue broker found at
>>> https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
>>> I modified the #include to look for zmq.h rather than zhelpers.h to get
>>> it to compile with ZMQ 3.x
>>>
>>> 2) a minimal server implementation that connects to the broker and does
>>> a zmq_recv. Code is:
>>> //
>>> //  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
>>> //  Connects to tcp://localhost:5560
>>> //
>>> #include <stdio.h>
>>> #include <zmq.h>
>>>
>>> int main (void)
>>> {
>>>       char buf[80];   // buffer for zmq_recv call
>>>       int len;        // len returned by zmq_recv
>>>       void *ctx;      // ZMQ context
>>>       void *sock;     // ZMQ socket
>>>
>>>       // init ZMQ context, create a REP socket, and connect to ROUTER
>>>       ctx = zmq_ctx_new();
>>>       sock = zmq_socket(ctx,ZMQ_REP);
>>>       zmq_connect (sock, "tcp://localhost:5560");
>>>
>>>       // do a basic zmq_recv call...warn if it fails
>>>       len = zmq_recv(sock,buf,sizeof(buf),0);
>>>       if (len < 0) {
>>>           printf("zmq_recv failed - %s\n",zmq_strerror(errno));
>>>       }
>>>
>>>       // Done. Clean up and end
>>>       zmq_close (sock);
>>>       zmq_term (ctx);
>>>       return 0;
>>> }
>>>
>>>
>>> Steps to repro are start server, start broker, kill broker (spec.
>>> ctrl-C), restart broker. Server aborts with the above assertion.
>>>
>>> Is there something my code should be doing differently?
>>>
>>> Thanks....Stuart
>>>
>>>
>>>
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev

From stu at compuserve.com  Tue Jul 31 23:19:10 2012
From: stu at compuserve.com (Stuart Brandt)
Date: Tue, 31 Jul 2012 17:19:10 -0400
Subject: [zeromq-dev] Assertion during zmq_recv in 3.x
In-Reply-To: <CALdZE=uvMBQjGcv2ZfGyY5dS9rQ17siXnC+Xe8e21RH4=j9WbQ@mail.gmail.com>
References: <5012EAB7.80500@compuserve.com>
	<CALdZE=sQJa3NKBc9-V+5zkfyOhA+EVNrm3je3YT646iTvx75SA@mail.gmail.com>
	<5017FE3E.9050701@compuserve.com>
	<CALdZE=uvMBQjGcv2ZfGyY5dS9rQ17siXnC+Xe8e21RH4=j9WbQ@mail.gmail.com>
Message-ID: <50184BCE.5020501@compuserve.com>

Checked out OK, so I closed it. Thanks for the quick response.

On 07/31/2012 03:26 PM, Martin Hurton wrote:
> Fixed in 9fab9937e516e78e6dbdf890e34f56cf80c9ff14. Can you check and
> close the issue if it fixes your problem.
>
> On Tue, Jul 31, 2012 at 5:48 PM, Stuart Brandt<stu at compuserve.com>  wrote:
>> Done. LIBZMQ-406
>>
>> Thanks for confirmation.  Let me know if there's any additional context
>> needed.
>>
>> - Stuart
>>
>> On 7/31/2012 6:04 AM, Martin Hurton wrote:
>>> Hi Stuart, could you please file an issue for this?
>>>
>>> - Martin
>>>
>>> On Fri, Jul 27, 2012 at 9:23 PM, Stuart Brandt<stu at compuserve.com>  wrote:
>>>> I get an assertion during a call to zmq_recv on 2 different OSes (OSX
>>>> 10.7.4 and CentOS 5.8) under two different builds of 3.x (v3.2.0-rc1 and
>>>> a recent build off latest at github.com/zeromq/libzmq).
>>>>
>>>> The assertion is:
>>>> Assertion failed: (msg_->flags ()&  msg_t::identity) == 0 (router.cpp:220)
>>>> Aborted
>>>>
>>>> The setup involves:
>>>> 1) the simple message queue broker found at
>>>> https://github.com/imatix/zguide/blob/master/examples/C/msgqueue.c
>>>> I modified the #include to look for zmq.h rather than zhelpers.h to get
>>>> it to compile with ZMQ 3.x
>>>>
>>>> 2) a minimal server implementation that connects to the broker and does
>>>> a zmq_recv. Code is:
>>>> //
>>>> //  Demo of assertion in ZMQ 3.x by REP socket connecting to ROUTER socket
>>>> //  Connects to tcp://localhost:5560
>>>> //
>>>> #include<stdio.h>
>>>> #include<zmq.h>
>>>>
>>>> int main (void)
>>>> {
>>>>        char buf[80];   // buffer for zmq_recv call
>>>>        int len;        // len returned by zmq_recv
>>>>        void *ctx;      // ZMQ context
>>>>        void *sock;     // ZMQ socket
>>>>
>>>>        // init ZMQ context, create a REP socket, and connect to ROUTER
>>>>        ctx = zmq_ctx_new();
>>>>        sock = zmq_socket(ctx,ZMQ_REP);
>>>>        zmq_connect (sock, "tcp://localhost:5560");
>>>>
>>>>        // do a basic zmq_recv call...warn if it fails
>>>>        len = zmq_recv(sock,buf,sizeof(buf),0);
>>>>        if (len<  0) {
>>>>            printf("zmq_recv failed - %s\n",zmq_strerror(errno));
>>>>        }
>>>>
>>>>        // Done. Clean up and end
>>>>        zmq_close (sock);
>>>>        zmq_term (ctx);
>>>>        return 0;
>>>> }
>>>>
>>>>
>>>> Steps to repro are start server, start broker, kill broker (spec.
>>>> ctrl-C), restart broker. Server aborts with the above assertion.
>>>>
>>>> Is there something my code should be doing differently?
>>>>
>>>> Thanks....Stuart
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> zeromq-dev mailing list
>>>> zeromq-dev at lists.zeromq.org
>>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>>> _______________________________________________
>>> zeromq-dev mailing list
>>> zeromq-dev at lists.zeromq.org
>>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev


From john_cerwin at yahoo.com  Tue Jul 31 23:52:20 2012
From: john_cerwin at yahoo.com (John Cerwin)
Date: Tue, 31 Jul 2012 14:52:20 -0700 (PDT)
Subject: [zeromq-dev] java.lang.UnsatisfiedLinkError: no jzmq in
	java.library.path
Message-ID: <1343771540.81840.YahooMailNeo@web113613.mail.gq1.yahoo.com>

Hi,

I am running Fedora 17 with ZMQ 2.2. It works fine under C, but I get the following error in Java when I execute the following command:

java -Djava.library.path=/usr/local/lib? -cp /home/john/jzmq/src/zmq.jar:.? local_lat tcp://127.0.0.1:5000 1 100

Exception in thread "main" java.lang.UnsatisfiedLinkError: no jzmq in java.library.path
??? at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1860)
??? at java.lang.Runtime.loadLibrary0(Runtime.java:845)
??? at java.lang.System.loadLibrary(System.java:1084)
??? at org.zeromq.ZMQ.<clinit>(ZMQ.java:34)
??? at local_lat.main(local_lat.java:36)


Here is what I have done so far:

1) Built zqm and jzmq on the same box (Fedora 17).
2) Followed the instructions here for Fedora: http://www.zeromq.org/bindings:java (***EXCEPT*** I am using zmq stable release 2.2 and NOT 2.1.10.)
3) Verified that libzmq is in /usr/local/lib, and that my java.library.path includes /usr/local/lib.?
4) Verfied that the classpath includes zmq.jar.?
5) Attempted to run it from both Eclipse and the shell? with the same resulting error.

It appears that the native library cannot be found.... but could it be the incorrect version of zmq (2.2 vs 2.1.10)?

Any help would be greatly appreciated!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120731/dec8e8a9/attachment.htm