[rabbitmq-discuss] Throughput observation with RabbitMQ-3.1.3 and Erlang R16B01 - Single Node and Cluster Node

Michael Klishin mklishin at gopivotal.com
Thu Aug 1 20:32:29 BST 2013

Priyanki Vashi:

> I have tried to with multiple producer and multiple consumer since my producer and consumer are single threaded (around 10 each) but it's the same statistics in test-1, which is around 5000 msg/sec. And since eventually we will use python clients in real application of rabbitmq, I think it would be good to know how it behaves with python clients ( in this case I chose python-pika).

As your other tests demonstrate, a single RabbitMQ node can go to 25K with your setup.

If a Python producer (or multiple instances of thee) demonstrate lower throughput, it's logical
to suspect that publishing is the bottleneck.

I haven't done client benchmarks in a while but 5-6 KGhz is about right for Pika on CPython
and also Bunny 0.9+ on CRuby. Java client and clients based on it easily go into the 20-25K

I can't offer any advice about possible throughput improvements with Pika but
it's a good idea to profile your producer to see what actually takes time.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20130801/5cc09188/attachment.pgp>

More information about the rabbitmq-discuss mailing list