[zeromq-dev] embedding zeromq question

Joshua Foster jhawk28 at gmail.com
Tue Jun 5 06:54:35 CEST 2012


The low-latency nature of ZeroMQ makes it ideal for embedded platforms. 
As of 3.x, ZeroMQ has no external dependencies (2.1 needs libuuid). My 
Windows version is 180KB, and my OSX version is 380KB. Your runtime 
needs will depend on how large the messages are and the number of 
messages expected in the queue(s).

Your benefit for 4-5 devices is a highly optimized, stable messaging 
solution that you don't need to write. The main features are messaging 
patterns (request/reply, publish/subscribe, and pipeline), 
re-connection, and atomic message delivery. I have found that I can 
develop applications much faster once I got over the initial learning 
curve. I would recommend reading at least the first chapter of the guide 
(http://zguide.zeromq.org/page:all#toc0).

Joshua

On 6/4/2012 8:29 PM, Ishigo, Kelvin K wrote:
> Hello,
> We are looking at distributed message queue solutions and zeromq looks promising... but we deploy on an embedded linux platform(s) (2.6+) which is hosted to several different processor architectures.
> Our main question is whether zeromq is a suitable in this type of environment where we are code footprint limited due to flash on the box.
> Also, is it overkill since we are basically planning to "share" amongst a limited number of devices (4-5)?
> Thank you.
>
> If there are existing threads, please do point me to them as I do not wish to rehash existing discussions.
>
> Kelvin
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev at lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev




More information about the zeromq-dev mailing list