Check behavior of Celery worker when Redis is down for a period of time

[stevenbal]

Product versie / Product version

Omschrijf het probleem / Describe the bug

Taiga ref Dimpact SSC 384

@alextreme suggested to investigate the behavior of the Celery worker when Redis is unavailable for some amount of time. This issue is only to investigate what happens, not to fix it yet, because we need to request approval for that. The time to investigate can be logged under Dimpact - PodiumD support.

A similar issue was resolved last year (#128), where Celery would not reconnect automatically when Redis was restarted. However, something similar seems to be happening again, perhaps when Redis is down for a longer amount of time?

To check:

• Does Celery reconnect automatically when Redis is restarted?
• Does Celery reconnect automatically when Redis is down for a prolonged amount of time (e.g. 5 minutes, 30 minutes, 1 hour)? Perhaps Celery stops retrying after some amount of time

Stappen om te reproduceren / Steps to reproduce

No response

Verwacht gedrag / Expected behavior

No response

[TimdeBeer1]

I tested this with the Objects-api. When Celery loses connection to Redis, it retries connecting. If Redis is restarted in time the worker reconnects successfully. But if the connection is down for an longer period (around 50 minutes), the celery worker gives a unrecoverable error, it stops running and needs to be restarted manually.

{"event": "Unrecoverable error: OperationalError('Error -2 connecting to redis:6379. Name or service not known.')", "exc_info": ["<class 'kombu.exceptions.OperationalError'>", "OperationalError('Error -2 connecting to redis:6379. Name or service not known.')", "<traceback object at 0x7f99207da240>"], "source": "app", "timestamp": "2026-01-30T09:20:56.267532Z", "logger": "celery.worker", "level": "critical"}