[rabbitmq-discuss] After weekly log rotatation was done, the command "rabbitmqctl rotate-logs" never stop
matthias at rabbitmq.com
Wed May 15 10:22:26 BST 2013
On 15/05/13 09:51, Nate wrote:
> 1. Did logs get rotated successfully in the past or did all log
> rotations get stuck?
> Only two of nodes in the cluster get stuck. And two clusters are in the
> same situation.
My question was whether the logs on these nodes ever got rotated
successfully in the past.
> 2. Your initial report showed a few more processes, including a
> beam.smpprocess. Presumably that's still there too, right? If so, please
> postthe complete command line for that.
> rabbitmq 8972 0.0 2.8 580024 510588 ? Ssl May12 0:00
> /usr/lib/erlang/erts-5.9/bin/beam.smp -- -root /usr/lib/erlang -progname
> erl -- -home /var/lib/rabbitmq -- -pa
> /usr/lib/rabbitmq/lib/rabbitmq_server-3.0.1/sbin/../ebin -noshell
> -noinput -hidden -sname rabbitmqctl8972 -s rabbit_control_main -nodename
> rabbit at rabbitmq1 -extra rotate_logs
> 3. I wanna know the root cause too! Therefore, I prefer to keeping the
> process and let it telling us the reason.
Unfortunately the rabbitmqctl node is a hidden node, so we cannot
connect to it. The best we can do is make it crash with
kill -SIGUSR1 8972
That should produce an erl_crash.dump. Not sure where it will end up
though. Possibly in the rabbitmq home dir. Anyway, I'm sure you'll
manage to track it down. Stick that crash dump on something like dropbox
and post a link.
More information about the rabbitmq-discuss