Redis connection reset occurs during unacked message mutex acquisition in Celery
warningConnection ManagementUpdated Mar 16, 2026(via Exa)
How to detect:
Connection reset error specifically occurs when Celery attempts to acquire a mutex lock for restoring visible unacked messages. The error happens in the kombu.transport.redis.Mutex context manager during the maybe_restore_messages timer operation, preventing proper message acknowledgment recovery.
Recommended action:
Monitor for ConnectionResetError stack traces showing kombu/transport/redis.py in restore_visible and Mutex functions. This indicates the connection failed during message restoration attempts. Check Redis connection stability during high-frequency operations. Review unacked message handling and consider adjusting message visibility timeout settings if connection resets correlate with long-running tasks.