Lookout service sends traffic to the Lookout DB Pruner when its running

[Sovietaced]

It appears that the Lookout DB pruner uses similar labels as the Lookout deployment and as a result traffic targeting the lookout servers gets split between the actual lookout pods and the lookout DB pruner pods. This caused a production issue for our retry service which queries lookout for failed jobs.

I believe this also causes issues with deletion since this uses the same reference owner as the other lookout related resources.

Should have a fix for this one soon. Will also need to audit the scheduler.