pltbkd opened a new pull request, #20733: URL: https://github.com/apache/flink/pull/20733
## What is the purpose of the change In [FLINK-28588](https://issues.apache.org/jira/browse/FLINK-28588), MetricStore supports multiple attempts of a subtask. However, `SubtaskMetricStore` doesn't have a clean mechanism. In failover scenario, `attempts` pile up and cause OOM. This PR fixes the issue by cleaning up the SubtaskMetricStore when JobDetails are updated. ## Brief change log - Enrich JobDetails.currentExecutionAttempts with all current execution attempts for each subtask, and make the first of the attempt list to be the representative attempt. The currentExecutionAttempts now contains all subtasks no matter whether the subtask contains more than one current attempt, so it is excluded from the jsonized JobDetails. -SubtaskMetricStore retains metrics of the potentially currentExecutionAttempts according to the JobDetails when updating. - MetricStore.currentExecutionAttempts is now renamed as representativeAttempts. ## Verifying this change This change added tests. ## Does this pull request potentially affect one of the following parts: - Dependencies (does it add or upgrade a dependency): no - The public API, i.e., is any changed class annotated with `@Public(Evolving)`: no - The serializers: no - The runtime per-record code paths (performance sensitive): no - Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: no - The S3 file system connector: no ## Documentation - Does this pull request introduce a new feature? no - If yes, how is the feature documented? not applicable -- 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]
