Possible OutOfRangeException when reading the stream log table after resuming from a long interruption

[zuston]

Search before asking

• I searched in the issues and found nothing similar.

Fluss version

0.9.0 (latest release)

Please describe the bug 🐞

When Flink recovers from a checkpoint after a long interruption, subtasks may throw an OutOfRangeException, as shown in the logs below.

I understand this is caused by local tablet logs being offloaded to remote storage, which makes the recovered split offsets stale and leads to an OutOfRange condition.
While this behavior is acceptable for Kafka-style logs, it is not appropriate for data lake–backed log tables. In this case, the job will keep failing indefinitely. We should handle this scenario explicitly to avoid permanent recovery failures.

Solution

No response

Are you willing to submit a PR?

• I'm willing to submit a PR!