You are viewing a plain text version of this content. The canonical link for it is here.
Posted to builds@flink.apache.org by GitBox <gi...@apache.org> on 2026/09/08 20:23:30 UTC

[GH] (flink-connector-kafka/fix-idle-reader-no-more-splits): Workflow run "Flink Connector Kafka CI" is working again!

The GitHub Actions job "Flink Connector Kafka CI" on flink-connector-kafka.git/fix-idle-reader-no-more-splits has succeeded.
Run started by GitHub user slachiewicz (triggered by slachiewicz).

Head commit for run:
4eded31c05b5c2fa8cbbad01796f49ecc4bd2ee5 / Sylwester Lachiewicz <sl...@apache.org>
[FLINK-40362] Fix idle reader hang when no-more-splits precedes metadata update

An idle reader (subtask with no assigned splits) that receives the no-more-splits
signal before the MetadataUpdateEvent never forwards it to the sub-readers created
by that update: the re-notification in handleSourceEvents was guarded by
!pendingSplits.isEmpty(), which is false in exactly this case. The reader then
returns NOTHING_AVAILABLE indefinitely and a bounded job never finishes.

Re-deliver the no-more-splits signal after processing a metadata update regardless
of whether pending splits exist. notifyNoMoreSplits() is idempotent and only fires
when the signal has already arrived, so this adds no new cross-component traffic.

Verified: DynamicKafkaSourceReaderTest#testIdleReaderFinishesWhenNoMoreSplitsArrivesBeforeMetadata
-> fails on main (NOTHING_AVAILABLE), passes with this change.

Generated-by: Claude Fable 5

Report URL: https://github.com/apache/flink-connector-kafka/actions/runs/34263930957

With regards,
GitHub Actions via GitBox