Partman stops all management operations if a table doesn't exist

[pgporada]

I detected this after running the alter table and partman startup command in #46.

-- The job identifier is 496805.
May 20 19:55:17 mydb partition-manager[642828]: 2022-05-20 19:55:17,509 - all_configured_tables_are_compatible - ERROR - Cannot proceed: Table orderToAuthz2_new_20220419 ['Unable to read information for orderToAuthz2_new_20220419']
May 20 19:55:17 mydb systemd[1]: partition-manager.service: Succeeded.
-- Subject: Unit succeeded

The partman error text output leads me to believe that partman will fail to perform actually useful operations since this table no longer exists in my database. This could lead to a table overrun issues and we'd have to re-migrate tables again.

[jcjones]

all_configured_tables_are_compatible is only checked for the migrate command, not for maintain, so it should only affect operator-hands-on operations, not automated ones. Still, we should fix this.