Add timeout monitoring to zlonk

Add better observability to zlonk, as the failure we experienced last week went unnoticed until the Data Team reported a failure on data ingestion.

• ensure we detect when the database fails to go into production mode after recovery
• add timeout monitoring (runbook)