We seemed to have run out allocated memory for repairing any database issues. That's been fixed. There should be no more interruptions!
