Dear backports team,
I am running rabbitmq-server 3.6.10-1~bpo9+1 and lately I have run into
a problem where while trying to restart a node (as part of a cluster)
after rebooting it. I noticed that the (idle) machine was waiting
indefinitely for Mnesia tables to become available, not rejoining the
cluster:
% sudo tail /var/log/rabbitmq/rabbit*log
[...]
=WARNING REPORT==== 30-Nov-2018::15:18:52 ===
Error while waiting for Mnesia tables: {timeout_waiting_for_tables,
[rabbit_user,rabbit_user_permission,
rabbit_vhost,rabbit_durable_route,
rabbit_durable_exchange,
rabbit_runtime_parameters,
rabbit_durable_queue]}
=INFO REPORT==== 30-Nov-2018::15:18:52 ===
Waiting for Mnesia tables for 600000 ms, 12 retries left
[..]
However, this struck me as interesting:
% sudo tail /var/log/rabbitmq/startup_err
systemd unit for activation check: "rabbitmq-server.service"
Failed to start socat error:enoent
systemd READY notification failed, beware of timeouts
After manually installing socat and restarting the service, the node
became available in the cluster almost immediately.
I'm wondering whether the rabbitmq-server version in backports is
missing a dependency on socat? I'm asking because the current version in
testing (3.7.8) indeed does have a dependency on socat...
Cheers
Sascha
Attachment:
signature.asc
Description: OpenPGP digital signature