Comment 11 for bug 1818260

UPDATE:

While I assumed this would be about the cluster-partition-handling it seems rather there was code introduced a few years ago [0] that attempts to remove nodes it is unaware of. This is almost certainly where the bug lies.

[0] https://github.com/openstack/charm-rabbitmq-server/commit/08b10513c5725fb740382668c47fc769a6f2936c