[rabbitmq-discuss] rabbitmq-server crashes hard while consuming 31GB of RAM

[Matthias Radestock]

Muharem,

On 21/12/11 10:19, Muharem Hrnjadovic wrote:
> Just before getting into the stuck state I ran "rabbitmqctl report"
> (please see https://bugs.launchpad.net/openquake/+bug/894024/comments/6)
> and the RAM usage command ypou kindly supplied (please see
> https://bugs.launchpad.net/openquake/+bug/894024/comments/7).

The latter report doesn't show all that much memory being used by the 
top 30 processes.

Btw, as of 2.7.1 you should be able to produce that report with

rabbitmqctl eval
'[{Mem, Pid, process_info(Pid)} || {Mem, Pid} <- 
lists:sublist(lists:reverse(lists:sort([{process_info(Pid, memory), Pid} 
|| Pid <- processes()])), 30)].'

(this is tweaked slightly to be more efficient when there are lots of 
processes, which I suspect is the case in your rabbit)

Try doing that *while* it is wedged. And also send us the output of 
'rabbitmqctl report'. Again, while rabbit is wedged.

My hunch is that you probably have many thousands of celery result queue 
processes.

Regards,

Matthias.