Celery SQS backend will stop consuming tasks after failures

The Fix

pip install celery==4.4.0rc5

Based on closed celery/celery issue #5844 · PR/commit linked

Production note: This usually shows up under retries/timeouts. Treat it as a side-effect risk until you can verify behavior with a canary + real traffic.

Jump to Verify Open PR/Commit

@@ -510,6 +510,10 @@ def on_failure(self, exc_info, send_failed_event=True, return_ok=False):
             elif ack:
                 self.acknowledge()
+            else:
+                # supporting the behaviour where a task failed and
+                # need to be removed from prefetched local queue

repro.py

from celery import Celery
app.conf.worker_prefetch_multiplier = 1

@app.task(base=BaseAsyncJobTask, acks_late=True, acks_on_failure_or_timeout=False)
def dead_letter_q_task():
    return 1/0

dead_letter_q_task.delay()
dead_letter_q_task.delay()
dead_letter_q_task.delay()

verify

Re-run the minimal reproduction on your broken version, then apply the fix and re-run.

fix.md

Option A — Upgrade to fixed release\npip install celery==4.4.0rc5\nWhen NOT to use: Do not use this fix if your application relies on tasks being retried after failure.\n\n

Why This Fix Works in Production

Trigger: SQS backend will stop consuming tasks after failures
Mechanism: The SQS backend did not reject tasks on failure, causing the worker to stop consuming tasks
Why the fix works: The SQS backend was modified to reject tasks on failure, preventing the worker from stopping after a failure. (first fixed release: 4.4.0rc5).

Production impact:

If left unfixed, the same config can fail only in production (env differences), causing startup failures or partial feature outages.

Why This Breaks in Prod

The SQS backend did not reject tasks on failure, causing the worker to stop consuming tasks
Production symptom (often without a traceback): SQS backend will stop consuming tasks after failures

Proof / Evidence

GitHub issue: #5844
Fix PR: https://github.com/celery/celery/pull/5843
First fixed release: 4.4.0rc5
Reproduced locally: No (not executed)
Last verified: 2026-02-09
Confidence: 0.85
Did this fix it?: Yes (upstream fix exists)
Own content ratio: 0.60

Discussion

High-signal excerpts from the issue thread (symptoms, repros, edge-cases).

Jump to Sources Open on GitHub

“can you send a PR with a test? did you try celery==4.4.0rc4?”

@auvipy · 2019-11-21 · source

“Tried with celery==4.4.0rc4, reproduced test case: https://github.com/galCohen88/kombu/pull/1/files”

@galCohen88 · 2019-11-21 · source

“Figured it out PR: https://github.com/celery/celery/pull/5843”

@galCohen88 · 2019-11-22 · source

“I would need a little guidance, as I'm not sure how I can access QoS attributes (https://github.com/celery/kombu/blob/master/kombu/transport/virtual/base.py#L182) from Task context”

@galCohen88 · 2019-11-21 · source

Failure Signature (Search String)

SQS backend will stop consuming tasks after failures
- [X] I have included all related issues and possible duplicate issues

Copy-friendly signature

signature.txt

Failure Signature
-----------------
SQS backend will stop consuming tasks after failures
- [X] I have included all related issues and possible duplicate issues

Error Message

Signature-only (no traceback captured)

error.txt

Error Message
-------------
SQS backend will stop consuming tasks after failures
- [X] I have included all related issues and possible duplicate issues

Minimal Reproduction

repro.py

from celery import Celery
app.conf.worker_prefetch_multiplier = 1

@app.task(base=BaseAsyncJobTask, acks_late=True, acks_on_failure_or_timeout=False)
def dead_letter_q_task():
    return 1/0

dead_letter_q_task.delay()
dead_letter_q_task.delay()
dead_letter_q_task.delay()

What Broke

Workers stopped consuming tasks after a failure, leading to task backlog.

Why It Broke

The SQS backend did not reject tasks on failure, causing the worker to stop consuming tasks

Fix Options (Details)

Option A — Upgrade to fixed release Safe default (recommended)

pip install celery==4.4.0rc5

When NOT to use: Do not use this fix if your application relies on tasks being retried after failure.

Use when you can deploy the upstream fix. It is usually lower-risk than long-lived workarounds.

Fix reference: https://github.com/celery/celery/pull/5843

First fixed release: 4.4.0rc5

Last verified: 2026-02-09. Validate in your environment.

When NOT to Use This Fix

Do not use this fix if your application relies on tasks being retried after failure.

Verify Fix

verify

Re-run the minimal reproduction on your broken version, then apply the fix and re-run.

Did This Fix Work in Your Case?

Quick signal helps us prioritize which fixes to verify and improve.

Prevention

Make timeouts explicit and test them (unit + integration) to avoid silent behavior changes.
Instrument retries (attempt count + reason) and alert on spikes to catch dependency slowdowns.

Version Compatibility Table

Version	Status
4.4.0rc5	Fixed

Related Issues

No related fixes found.

Cluster: celery:configuration-error Celery hub Celery best practices All hubs All clusters

Related clusters: Data consistency Duplicates Duplicate processing

Sources

We don’t republish the full GitHub discussion text. Use the links above for context.

Celery SQS backend will stop consuming tasks after failures (Fix)