Comment 7 for bug 1881735

I think I may have hit this.

In a test with recently saw mysql-innodb-cluster fail to cluster, hanging in "waiting/Not all instances clustered" status, respectively in "waiting/Instance not yet in the cluster"

This seems intermittent, I only saw this once among abt. 4 test runs.

Checking logs of one unit I found this error:

-- snip --
2021-01-28 21:03:38 ERROR juju-log Failed adding instance 10.0.8.235 to cluster: Cannot set LC_ALL to locale en_US.UTF-8: No such file or directory

Clone based recovery selected through the recoveryMethod option

ESC[36mNOTE: ESC[0mGroup Replication will communicate with other members using '10.0.8.235:33061'. Use the localAddress option to override.
...
Adding instance to the cluster...
Traceback (most recent call last):
  File "<string>", line 3, in <module>
SystemError: RuntimeError: Cluster.add_instance: RESET MASTER is not allowed because Group Replication is running.

-- snip --

Charm version:

cs:~openstack-charmers-next/mysql-innodb-cluster-61

LXD container running focal

Additional diagnostics: https://private-fileshare.canonical.com/~sabaini/e8a3afda-c460-4e61-9d33-135cfc666c0f/juju-crashdump-812b5deb-7fd1-4dde-bdd4-4a37b996c214.tar.xz