Querki shouldn't have so many tables hanging around
Summary: Querki originally simulated Cassandra partitions by using Lots and Lots of MySQL tables. Those are mostly no longer needed, and they slow down DB upgrades a lot.
Need to analyze this, figure out which ones are no longer needed, and come up with a nice safe strategy to get rid of them after archiving them.
This will be an ongoing project for a while, as more functionality moves to Cassandra.