[rabbitmq-discuss] RabbitMQ keeps crashing, wrong rabbitmqctl nodedown

David Wragg david at rabbitmq.com
Wed Sep 15 12:28:12 BST 2010


Hi Josua,

"josua.troesch at bluewin.ch" <josua.troesch at bluewin.ch> writes:
> Our RabbitMQ instance keeps crashing with an error like this:
> eheap_alloc: Cannot allocate 373662860 bytes of 
> memory (of type "old_heap")

This indicates that rabbit is running out of memory.  

The normal approach to this kind of problem is to use rabbitmqctl to
find out where the messages are that are taking up your memory.

> Its still a version 1.6.0 but because of some application constraints
> we can't update yet. 

1.6.0 is really really old, and the handling of memory constraints has
been greatly improved since then.  What kind of application constraints
are preventing you upgrading?

> Also, I can't check the status with rabbitmqctl as this always reports
> a {badrpc,nodedown}

Even before the server crash (or does the crash happen when you start
rabbitmq-server)?

David

-- 
David Wragg
Staff Engineer, RabbitMQ
SpringSource, a division of VMware


More information about the rabbitmq-discuss mailing list