[Openstack-operators] Services last time not updated and in XXX state

Jesse Pretorius jesse.pretorius at gmail.com
Thu Oct 17 10:15:56 UTC 2013


On 17 October 2013 11:13, Robert van Leeuwen <
Robert.vanLeeuwen at spilgames.com> wrote:

>  It was rabbitMQ.
> Apparently if Rabbit thinks it has not enough potential free space Rabbit
> will just stop accepting messages for the queue.
> To bad it did not report any of this in the Rabbit logfile :(
>
>
Ah yes. I've seen this more and more lately. You need to make sure that
RabbitMQ has enough memory to operate. It has a default threshold of 40%
memory usage as a cap before it blocks further messages until the memory
usage is reduced enough for it to continue.

Some tips:
1. Ensure that your RabbitMQ servers have enough memory
2. Monitor the memory usage to determine your baseline
3. Implement the management plugin so that you can quickly and easily check
on the health and status of the service and queues
4.  Upgrade RabbitMQ to the latest available to resolve a variety of memory
leaks and improve the monitoring capabilities
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20131017/499fc9fa/attachment.html>


More information about the OpenStack-operators mailing list