Terminating celery-flower pod caused old tasks to be re-queued

[jrochette]

Describe the bug
Running celery-flower in a kubernetes cluster. Restarting the pod with a kubectl delete pod ... command caused many old completed tasks to be re-queued. Maybe it's the expected behaviour? I'm trying to understand why this happens. I'm running celery with a Redis broker and the following settings

CELERY_TASK_ACKS_LATE = True
CELERY_WORKER_PREFETCH_MULTIPLIER = 1

To Reproduce

• Run celery flower in k8s using the mher/flower:1.2.0 image and celery flower --purge_offline_workers=900--persistent=True --db=/flower/flower
• Delete the pod kubectl delete pod celery-flower

Expected behavior
I would not expect any tasks to be re-queued upon terminating flower

System information
python -c 'from flower.utils import bugreport; print(bugreport())'

flower   -> flower:1.2.0 tornado:6.2 humanize:4.3.0
software -> celery:5.2.7 (dawn-chorus) kombu:5.2.4 py:3.10.6
            billiard:3.6.4.0 redis:4.3.4
platform -> system:Linux arch:64bit
            kernel version:5.10.225-213.878.amzn2.x86_64 imp:CPython
loader   -> celery.loaders.app.AppLoader
settings -> transport:rediss results:disabled

deprecated_settings: None