luoyuxia opened a new pull request, #3355: URL: https://github.com/apache/fluss/pull/3355
### Purpose Linked issue: close #3354 Preserve the original log-only startup semantics after a Flink source restore. When the source starts from a specified log position, it should not initialize lake snapshot splits after restore just because `lakeSource` is available again. ### Brief change log - Persist an empty `remainingHybridLakeFlussSplits` list when the enumerator checkpoints with `lakeSource == null`. - Reuse the existing non-null pending hybrid split sentinel so restored enumerators skip lake snapshot generation without changing the enumerator state format. - Add a unit test covering the checkpoint state produced by a log-only source enumerator. ### Tests - `./mvnw -pl fluss-flink/fluss-flink-common -Dtest=FlinkSourceEnumeratorTest#testLogOnlySourceCheckpointsEmptyHybridLakeFlussSplits test` - `git diff --check` ### API and Format No public API, storage format, or checkpoint serializer version change. The fix only changes the value stored in the existing enumerator state field for log-only source checkpoints. ### Documentation No new feature or user-facing configuration change. ### Generative AI disclosure - [x] Yes: OpenAI Codex was used to author this PR following the repository AGENTS.md guidance. -- 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]
