sjyangkevin opened a new pull request, #63454:
URL: https://github.com/apache/airflow/pull/63454
<!-- SPDX-License-Identifier: Apache-2.0
https://www.apache.org/licenses/LICENSE-2.0 -->
<!--
Thank you for contributing!
Please provide above a brief description of the changes made in this pull
request.
Write a good git commit message following this guide:
http://chris.beams.io/posts/git-commit/
Please make sure that your code changes are covered with tests.
And in case of new features or big changes remember to adjust the
documentation.
Feel free to ping (in general) for the review if you do not see reaction for
a few days
(72 Hours is the minimum reaction time you can expect from volunteers) - we
sometimes miss notifications.
In case of an existing issue, reference it using one of the following:
* closes: #ISSUE
* related: #ISSUE
-->
related: #62887
### Summary
This PR implements executor callback support for KubernetesExecutor,
continuing the work started in #61153 and #62645. Synchronous callbacks are
routed to an executor to run as workloads.
### Approach
Callbacks run as Kubernetes pods, the same isolation model used for tasks.
The workload (an ExecuteCallback JSON payload) is passed to the
`execute_workload` entrypoint that tasks use, and the pod's exit code
determines success or failure.
The implementation threads callbacks through the existing executor pipeline
with annotation-based discrimination to distinguish callback pods from task
pods:
- Callback pods carry a `callback_id` annotation instead of
`dag_id`/`task_id`/`try_number`
- The watcher branches on this annotation to extract the right key type
- `_change_state()` dispatches on key type (str = callback, TaskInstanceKey
= task)
Callback pod lifecycle
```
Scheduler
└── _process_workloads(ExecuteCallback)
└── task_queue.put(KubernetesJob(callback_key, [workload], None,
None))
AirflowKubernetesScheduler.run_next()
└── _run_next_callback()
└── pod: python -m airflow.sdk.execution_time.execute_workload
--json-string <ExecuteCallback JSON>
annotations: {callback_id: <uuid>}
labels: {airflow-worker: <job_id>}
KubernetesJobWatcher (sees callback_id annotation → passes CallbackKey)
└── result_queue.put(KubernetesResults(callback_key, state, ...))
KubernetesExecutor._change_state()
└── isinstance(key, str) → _change_callback_state()
└── event_buffer[key] = CallbackState.SUCCESS / FAILED
```
Currently `execute_workload.py` calls `execute_callback_workload()` inline
in the pod process. Once `supervise_callback()` from (#62645) is merged, need
to update `_execute_callback()` in `execute_workload.py`
### Follow-ups
- Pod adoption on scheduler restart: Orphaned callback pods after a
scheduler restart are not currently adopted (unlike task pods).
- Callback revocation: No revoke_callback() support yet — callbacks can't be
cancelled mid-flight.
- supervise_callback() integration: Pending #62645; see above.
---
##### Was generative AI tooling used to co-author this PR?
<!--
If generative AI tooling has been used in the process of authoring this PR,
please
change below checkbox to `[X]` followed by the name of the tool, uncomment
the "Generated-by".
-->
- [X] Yes (please specify the tool below)
Generated-by: [Claude Code (claude-opus-4-6)] following [the
guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions)
---
* Read the **[Pull Request
Guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#pull-request-guidelines)**
for more information. Note: commit author/co-author name and email in commits
become permanently public when merged.
* For fundamental code changes, an Airflow Improvement Proposal
([AIP](https://cwiki.apache.org/confluence/display/AIRFLOW/Airflow+Improvement+Proposals))
is needed.
* When adding dependency, check compliance with the [ASF 3rd Party License
Policy](https://www.apache.org/legal/resolved.html#category-x).
* For significant user-facing changes create newsfragment:
`{pr_number}.significant.rst`, in
[airflow-core/newsfragments](https://github.com/apache/airflow/tree/main/airflow-core/newsfragments).
You can add this file in a follow-up commit after the PR is created so you
know the PR number.
--
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]
