Improvements to the storage_migration.rb script

A few improvements have been proposed:

Auto log to a file without the need for tee
Dynamic wait time
Better Error handling
One example, we query all projects at the start, and then query each project id later, if the project doesn't exist that second time, the script bails
Monitor for alerts for each file server being operated on to adjust whether more work should be queued
Detect job failures so we aren't waiting unnecessarily