Comment 9 for bug 1833730

Looking at the logs, controller-0 came up at 2019-06-21T12:59:18. We see mariadb coming up initially at 13:07:41 but it fails to join the DB cluster and aborts at 13:07:51. We see another attempt at 13:08:44 and it is synced with the DB cluster at 13:09:08.

This doesn't explain why the other mariadb server isn't handling the requests though, and we need the logs from the other controller node to answer that.