Event Details:
On December 10th, 2021 between 21:46 UTC and 22:36 UTC, the mail processing system at Area 1 experienced sporadic mail delays in two processing regions, though all protection and email delivery remained intact and fully functional.
Root Cause & Actions
The delay was due to our link analysis subsystem evaluating too many links simultaneously. During the incident the average time spent in queue before messages were processed, normally less than 1 second, rose to an average of 8 minutes in the affected regions. All messages were successfully queued and we did not observe delays in messages being accepted and entering the queued state. No messages were lost during the event.
The Area 1 engineering team has developed and deployed added measures to ensure our link analysis system does not cause message processing delays in the future.