The performance degradation was caused by the initiation of index maintenance on the database for our US-01 application stack. Stopping the index maintenance resulted in normal performance levels. We have created additional logic to prevent re-indexing during normal business hours or periods of expected heavy use.