At approximately 11:45 AM EST we were forced to restart our primary database clusters for an urgent performance issue that could not wait until after hours.
If you received "504 Gatweay Timeout" error from one of our web services it had to do with the service restart. As of 11:52 AM everything is back to normal.