[zeromq-dev] epgm performance numbers

Rohan Bedarkar r_bedarkar at yahoo.com
Thu Oct 25 19:04:05 CEST 2012

Unfortunately no luck. Just tried varying the multicast rates on both sides anywhere between 1000 to 10000 to even 1000000 but no change. Its a little counterintuitive that lowering rates can help especially if HWM is sufficiently high. I am using 32 bit SuSe on one side and 64 bit SuSe on the other side. 

Are there any configs that I might have overlooked? 

Both my repeater application and originating application are very light so I doubt these latencies are anywhere in there. If I simply switch to TCP the performance is really good.

Rohan Bedarkar
MS, Computer Science 

University of Chicago
rohanb at cs.uchicago.edu


 From: Steven McCoy <steven.mccoy at miru.hk>
To: Rohan Bedarkar <r_bedarkar at yahoo.com>; ZeroMQ development list <zeromq-dev at lists.zeromq.org> 
Sent: Thursday, October 25, 2012 11:47 AM
Subject: Re: [zeromq-dev] epgm performance numbers

On 25 October 2012 12:28, Rohan Bedarkar <r_bedarkar at yahoo.com> wrote:

>Hi Steven
>I tried numbers like 10,000 and even then it took almost 7 seconds. 
>It seems the throughput is the bottleneck here.... Thoughts? Do I need PGM tweeking? 
> Transport:                     ZMQ
> Protocol:                      epgm
> Messages recd:                 10000
> Total Time:                    7930204 us   
> Throughput:                    0.15132 Mbps 
> Avg per msg:                   793.02 us   

This suggests you need to lower the rate limit until you find a level that your environment can sustain.

What platform is this test being executed on?  On Windows it is quite drastic the performance improvement by throttling outgoing packets as the IP stack is generally the bottleneck.

zeromq-dev mailing list
zeromq-dev at lists.zeromq.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.zeromq.org/pipermail/zeromq-dev/attachments/20121025/c31bab00/attachment.htm>

More information about the zeromq-dev mailing list