lozuwa opened a new issue, #27753:
URL: https://github.com/apache/airflow/issues/27753
### Apache Airflow version
Other Airflow 2 version (please specify below)
### What happened
Airflow 2.2.4
We have a task that runs a slack notification which usually takes 1 to 3
seconds to finish. However, we noticed it was stuck on queued state for some
hours. Below I provide some logs of the scheduler, there it can be seen the
task was correctly queued and sent to the worker. However, when it came back as
successful, it didn't update the task state. As can be seen in the latest log.
```bash
Nov 17, 2022 @ 13:02:16.203 | [2022-11-17 17:02:16,203]
{scheduler_job.py:570} INFO - TaskInstance Finished: dag_id=***, task_id=***,
run_id=scheduled__2022-11-17T16:00:00+00:00, run_start_date=None,
run_end_date=None, run_duration=None, state=queued, executor_state=failed,
try_number=1, max_tries=1, job_id=None, pool=default_pool, queue=default,
priority_weight=1, operator=SlackAPIPostOperator
-- | --
| Nov 17, 2022 @ 13:02:15.943 | [2022-11-17 17:02:15,942]
{scheduler_job.py:473} INFO - Sending TaskInstanceKey(dag_id='***',
task_id='***', run_id='scheduled__2022-11-17T16:00:00+00:00', try_number=1) to
executor with priority 1 and queue default
| Nov 17, 2022 @ 13:02:15.943 | [2022-11-17 17:02:15,943]
{base_executor.py:82} INFO - Adding to queue: ['airflow', 'tasks', 'run',
'***', '***', 'scheduled__2022-11-17T16:00:00+00:00', '--local', '--subdir',
'DAGS_FOLDER/***.py']
```
### What you think should happen instead
The task updates the status to completed and doesn't hang in queued state.
### How to reproduce
This happens randomly.
### Operating System
Official docker image
### Versions of Apache Airflow Providers
None, custom code.
### Deployment
Official Apache Airflow Helm Chart
### Deployment details
Helm, kubernetes, celery executor.
### Anything else
_No response_
### Are you willing to submit PR?
- [X] Yes I am willing to submit a PR!
### Code of Conduct
- [X] I agree to follow this project's [Code of
Conduct](https://github.com/apache/airflow/blob/main/CODE_OF_CONDUCT.md)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
