Description of problem:

When shutting down Satellite services there is a HUGE scroll of pulpcore-[content,api] Python 3.11 backtrace due to the loss of PostgreSQL connection.  This is... disconcerting to say the least.

How reproducible:

Any time Satellite services are stopped.  Even in the correct manner.

Is this issue a regression from an earlier version:

Seems to be.

Steps to Reproduce:

1. "satellite-maintain service stop"

2. Observe /var/log/messages

Actual behavior:

The log file gains a few hundred lines of moderately scary backtrace scroll, which surely isn't anything other than worrying to sysadmins.

Expected behavior:

The expected way to stop Satellite services shouldn't generate errors like this.

Business Impact / Additional info:

Reduced confidence in Satellite's correct functionality and data integrity (given it's to do with the back end DB).