Details
-
Bug
-
Status: Closed
-
Blocker
-
Resolution: Fixed
-
0.17.1
-
None
-
None
-
Reviewed
Description
I upgraded a big cluster, out of which 10 nodes did not get upgraded.
The namenode log showed excessive exceptions, causing the namenode log to ate the entire partition space, in this case close to 700GB log file was generated on the namenode.
Attachments
Attachments
Issue Links
- is duplicated by
-
HADOOP-3802 IPC - Heartbeat exceptions filling up log files
- Closed