abtqian opened a new pull request, #56445:
URL: https://github.com/apache/spark/pull/56445
<!--
Thanks for sending a pull request! Here are some tips for you:
1. If this is your first time, please read our contributor guidelines:
https://spark.apache.org/contributing.html
2. Ensure you have added or run the appropriate tests for your PR:
https://spark.apache.org/developer-tools.html
3. If the PR is unfinished, add '[WIP]' in your PR title, e.g.,
'[WIP][SPARK-XXXX] Your PR title ...'.
4. Be sure to keep the PR description updated to reflect all changes.
5. Please write your PR title to summarize what this PR proposes.
6. If possible, provide a concise example to reproduce the issue for a
faster review.
7. If you want to add a new configuration, please read the guideline first
for naming configurations in
'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
8. If you want to add or modify an error type or message, please read the
guideline first in
'common/utils/src/main/resources/error/README.md'.
-->
### What changes were proposed in this pull request?
<!--
Please clarify what changes you are proposing. The purpose of this section
is to outline the changes and how this PR fixes the issue.
If possible, please consider writing useful notes for better and faster
reviews in your PR. See the examples below.
1. If you refactor some codes with changing classes, showing the class
hierarchy will help reviewers.
2. If you fix some SQL features, you can provide some references of other
DBMSes.
3. If there is design documentation, please add the link.
4. If there is a discussion in the mailing list, please add the link.
-->
When a Spark driver container running on Kubernetes restarts (e.g., due to
OOM) without the driver pod being deleted, executor pods from the previous
driver lifetime are left in a terminal state (Failed or Completed). These
orphaned pods are not cleaned up by Kubernetes garbage collection because they
still hold an ownerReference to the original driver pod (which still exists),
and the restarted driver no longer tracks them.
This PR adds a cleanTerminalExecutorPodsOnStart() method in
KubernetesClusterSchedulerBackend that runs at driver startup — before the pod
allocator is started — to detect and delete any executor pods in Failed or
Completed phase that share the same app name label as the current application.
The cleanup logic:
1. Lists executor pods filtered by spark-role=executor and the app name
label
2. Filters for pods in terminal phases (Failed or Completed)
3. Deletes them via the Kubernetes client
4. Logs a warning and continues if the cleanup fails (non-fatal)
### Why are the changes needed?
<!--
Please clarify why the changes are needed. For instance,
1. If you propose a new API, clarify the use case for a new API.
2. If you fix a bug, you can clarify why it is a bug.
-->
When the driver container is restarted by Kubernetes (e.g., due to OOM
kill), the driver pod itself is not recreated — only the container inside it
restarts. As a result:
- The old executor pods remain with ownerReference pointing to the
still-existing driver pod, so Kubernetes garbage collection does not remove
them.
- The restarted driver has no knowledge of those old executors and will
spawn new ones, leaving the old ones as zombie pods in Failed/Completed state.
This causes resource leakage and, in long-running applications like Spark
Thrift Server, accumulates stale pods over time.
### Does this PR introduce _any_ user-facing change?
<!--
Note that it means *any* user-facing change including all aspects such as
new features, bug fixes, or other behavior changes. Documentation-only updates
are not considered user-facing changes.
If yes, please clarify the previous behavior and the change this PR proposes
- provide the console output, description and/or an example to show the
behavior difference if possible.
If possible, please also clarify if this is a user-facing change compared to
the released Spark versions or within the unreleased branches such as master.
If no, write 'No'.
-->
Yes. On driver restart, Spark will now automatically delete orphaned
executor pods in Failed or Completed state (matched by app name label) before
starting pod allocation. Users running Spark on Kubernetes — especially
long-lived services like Thrift Server — will see stale executor pods cleaned
up on each driver restart instead of accumulating indefinitely.
### How was this patch tested?
<!--
If tests were added, say they were added here. Please make sure to add some
test cases that check the changes thoroughly including negative and positive
cases if possible.
If it was tested in a way different from regular unit tests, please clarify
how you tested step by step, ideally copy and paste-able, so that other
reviewers can test and check, and descendants can verify in the future.
If tests were not added, please describe why they were not added and/or why
it was difficult to add.
If benchmark tests were added, please run the benchmarks in GitHub Actions
for the consistent environment, and the instructions could accord to:
https://spark.apache.org/developer-tools.html#github-workflow-benchmarks.
-->
A unit test SPARK-37856: cleanTerminalExecutorPodsOnStart deletes Failed
and Completed executor pods was added to
KubernetesClusterSchedulerBackendSuite. The test sets up mocks for executor
pods in Failed, Completed, and Running phases, calls start(), and asserts that
only the terminal pods are passed to
kubernetesClient.resourceList(...).delete() — the running pod is not included.
### Was this patch authored or co-authored using generative AI tooling?
<!--
If generative AI tooling has been used in the process of authoring this
patch, please include the
phrase: 'Generated-by: ' followed by the name of the tool and its version.
If no, write 'No'.
Please refer to the [ASF Generative Tooling
Guidance](https://www.apache.org/legal/generative-tooling.html) for details.
-->
Yes
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
