Details
-
Improvement
-
Status: Closed
-
Major
-
Resolution: Fixed
-
2.0.0-alpha
-
None
-
Reviewed
Description
IBRs from thousands of DNs under load will degrade NN performance due to excessive write-lock contention from multiple IPC handler threads. The IBR processing is quick, so the lock contention may be reduced by coalescing multiple IBRs into a single write-lock transaction. The handlers will also be freed up faster for other operations.
Attachments
Attachments
Issue Links
- breaks
-
HDFS-10641 TestBlockManager#testBlockReportQueueing fails intermittently
- Resolved
- is related to
-
HDFS-10563 Block reports could be silently dropped by NN
- Resolved
-
HDFS-12619 Do not catch and throw unchecked exceptions if IBRs fail to process
- Resolved
- relates to
-
HDFS-14186 blockreport storm slow down namenode restart seriously in large cluster
- Reopened
-
HDFS-13554 TestDatanodeRegistration#testForcedRegistration does not shut down cluster
- Resolved