Rubygem-staypuft: HA: All services are missing max_retries=-1 settings for accessing the DB.
This means that if the DB is not started before the service, the service will fail to start. This behaviour is currently mitigated by: start-failure-is-fatal: false, but it is technically incorrect.
set max_retries=-1 on all services with access to the DB
add a start order constrain for services on galera/mysql. This will guarantee that the services will be always started _after_ the db. In the manual how-to I use solution1 and it is known to work.
Also a maintanance to the DB will triger a chain effect of stopping all services, while that's not required with solution1.
I'm going with Solution 1 unless somebody objects.
[root@maca25400702875 ~]# grep -P '^max_retries' /etc/*/*conf
grep: /etc/httpd/conf: Is a directory
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.