[FLINK-37370] [Observer] Finished batch jobs throw ReconciliationException and never reach FINISHED in the CR #948
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ption and never reach FINISHED in the CR
gyfora
reviewed
Feb 22, 2025
...s-operator/src/main/java/org/apache/flink/kubernetes/operator/observer/SnapshotObserver.java
Outdated
Show resolved
Hide resolved
gyfora
requested changes
Feb 24, 2025
| }); | ||
| } catch (Exception e) { | ||
| if (ExceptionUtils.findThrowable(e, RestClientException.class) | ||
| .map(ex -> ex.getMessage().contains("Checkpointing has not been enabled")) |
Contributor
There was a problem hiding this comment.
Is message always not null here?
Contributor
Author
There was a problem hiding this comment.
Good point. I'm not sure. I'll add logic to check for null.
Comment on lines
450
to
469
| ctx.getFlinkService() | ||
| .getLastCheckpoint(JobID.fromHexString(jobId), ctx.getObserveConfig()) | ||
| .ifPresentOrElse( | ||
| snapshot -> jobStatus.setUpgradeSavepointPath(snapshot.getLocation()), | ||
| () -> { | ||
| if (ReconciliationUtils.isJobCancelled(status)) { | ||
| // For cancelled jobs the observed savepoint is always definite, | ||
| // so if empty we know the job doesn't have any | ||
| // checkpoints/savepoints | ||
| jobStatus.setUpgradeSavepointPath(null); | ||
| } | ||
| }); | ||
| } catch (Exception e) { | ||
| if (ExceptionUtils.findThrowable(e, RestClientException.class) | ||
| .map(ex -> ex.getMessage().contains("Checkpointing has not been enabled")) | ||
| .orElse(false)) { | ||
| LOG.warn( | ||
| "Checkpointing not enabled for job {}, skipping checkpoint observation", | ||
| jobId, | ||
| e); |
Contributor
There was a problem hiding this comment.
I wonder if the try/catch logic should be part of getLastCheckpoint of the flink service. That would mean that anywhere else we call this in the future we get a consistently good behaviour for batch jobs
Contributor
Author
There was a problem hiding this comment.
Yeah, I thought about that. I'm open to either. We would just have getLastCheckpoint catch the exception and return optional.empty()
Contributor
There was a problem hiding this comment.
I think that would make sense 👍
gyfora
approved these changes
Feb 25, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What is the purpose of the change
This pull request fixes a bug where in the observation phase, the operator tries to observe savepoint information for batch jobs and fails since checkpointing is not enabled for batch jobs.
More information in: https://issues.apache.org/jira/browse/FLINK-37370
Brief change log
Some questions
Verifying this change
If you think my approach is good, as a next step, happy to add some tests for the new logic I've introduced. Where would you like to see tests?
Does this pull request potentially affect one of the following parts:
CustomResourceDescriptors: noDocumentation