[rabbitmq-discuss] Feature Req / Bug list

Simon MacMullen simon at rabbitmq.com
Fri Oct 25 17:40:35 BST 2013


On 25/10/2013 5:13PM, Graeme N wrote:
> Err, your script invokes amqp-publish(1) in a loop. 100,000 times. I
>  suspect most of the slowness is due to the time taken to fork that
> many processes, open and close that many AMQP connections, and so
> on. I would guess this goes faster on an SSD because you can fork()
> faster.
>
>
> Hmmm... I wouldn't expect this to be the case, since I can fork
> 100000 echos is about 1s on this box. I am willing to admit that
> there could be issues in the way the data's being generated out of
> bash, or for any number of other reasons though.

It might be setting up and tearing down 100k AMQP connections then...
anyway, I ran it myself and it is several orders of magnitude slower
than it should be.

> Yeah, I did that intentionally in my test script so it'd be easier to
> reproduce the issue. I was actually doing this sequentially when I
> initially wrote my script to rebalance queues across the cluster, and
> still encountered the issue, it just took a lot more cluster
> operations before it happened. Presumably if you fix the parallel
> case, (by making the API handler block when running API reqs hits an
> upper bound or something?), it'll automatically solve any issues with
> the sequential case.

I should think so. But thanks, that's useful to know.

Cheers, Simon

-- 
Simon MacMullen
RabbitMQ, Pivotal


More information about the rabbitmq-discuss mailing list