Details
-
Bug
-
Status: Resolved
-
Critical
-
Resolution: Duplicate
-
2.0.5-alpha
-
None
-
None
Description
Currently, the AM writes the job-specific information to HDFS only after it finishes; the JHS needs this info to display anything. If the AM fails, this info is not written and the JHS fails to display anything for that job.
While JHS on top of AHS might address this issue, it would be nice to have a solution in the interim.
Attachments
Attachments
Issue Links
- depends upon
-
YARN-910 Allow auxiliary services to listen for container starts and completions
- Closed
- duplicates
-
MAPREDUCE-5641 History for failed Application Masters should be made available to the Job History Server
- Open
- is related to
-
MAPREDUCE-4428 A failed job is not available under job history if the job is killed right around the time job is notified as failed
- Open
-
MAPREDUCE-4559 Job logs not accessible through job history server for AM killed due to am.liveness-monitor expiry
- Open