Details
-
Bug
-
Status: Resolved
-
Critical
-
Resolution: Fixed
-
1.13.6, 1.14.3, 1.15.0
Description
The retryable cleanup does not work properly if there's an error appearing during the FileSystem.delete call which is used within FileStateHandle.discardState. Some FileSystem implementations (e.g. S3 presto; see PrestoS3FileSystem:512 through PrestoS3FileSystem.delete(Path, boolean)) return false in case of an error which will be swallowed in FileStateHandle.discardState.
Attachments
Issue Links
- Discovered while testing
-
FLINK-26388 Release Testing: Repeatable Cleanup (FLINK-25433)
- Resolved
- is caused by
-
FLINK-4910 Introduce safety net for closing file system streams
- Resolved
- is fixed by
-
FLINK-25484 TableRollingPolicy do not support inactivityInterval config which is supported in datastream api
- Closed
- relates to
-
FLINK-26488 FileSystem.listFiles is not implemented consistently
- Open
-
FLINK-26484 FileSystem.delete is not implemented consistently
- Resolved
- links to