Details
-
Bug
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
3.21.2
-
None
-
Unknown
Description
We have had the following scenario happen to us on a SQS consumer with greedy=true:
poll 1: fails with exception, causing error count to increase to 1 (due to sqs problem)
poll 2: successfully poll, and since greedy, continues with the next poll. This continued for 1 hour as the queue it polled from had many messages
Since the error count wasn't updated, then our readiness check remained DOWN for a full hour even though the application actually processed messages of the queue.