New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Crash when trying to reload rabbitmq-management plugin after removing a node #298
Comments
I edited out 90% of the log because it had the same exception for every channel. If stats DB cannot be reached, channels won't be able to emit stats to it. There is only one node that collects statistics and it is entirely in RAM — all of the data it has is transient by design. Enabling and disabling the plugin is unnecessary — there's plenty of threads on the mailing list that demonstrate how to make stats DB terminate with It looks like the stats DB became unreachable right after it queried a bunch of channels. I'm not sure how much can be improved here. |
Couple of things:
|
|
@evanccnyc there should be a I'll add a note to the site, which is open source, by the way :) |
Thanks. Ah next time I just submit a pull request for that. |
That stacktrace is very interesting in itself. I've seen similar one, and I can't grasp what's happening. And there are 2 things that I can't explain:
Everything starts with mnesia event about inconsistent database.
which it doesn't know how to handle (as there is no even remotely similar clause of And it results in an even more strange stacktrace:
How did And on another occasion I've seen queue slave process terminated with
What have queue slave process and DTLS (which is not even used anywhere) in common? What is responsible for all those |
Hi there,
I ran into a rabbitmq crash after the following scenario:
RabbitMQ 3.5.3
Ubuntu 14.04
3 servers, 1c,1d,1e
There was a network blip and 1e briefly connected/reconnected, at which point the management GUI showed it not connected to the cluster. We drained the queues and then disconnected the 1e node. At which point we lost all information on the management GUI on the remaining two servers. The interface showed lots of ????'s and error messages like "Statistics database could not be contacted.".
At which point we removed the management plugin from one of the servers (1c) via the command line and tried to readd it back with: "rabbitmq-plugins enable rabbitmq_management"
At which point 1c crashed with the following debug message:
The text was updated successfully, but these errors were encountered: