Details
-
Bug
-
Status: Resolved
-
Critical
-
Resolution: Fixed
-
2.6.1
-
None
-
Reviewed
Description
When NameNode is busy a DataNode can timeout sending a block report. Then it sends the block report again. Then NameNode while process these two reports at the same time can interleave processing storages from different reports. This screws up the blockReportId field, which makes NameNode think that some storages are zombie. Replicas from zombie storages are immediately removed, causing missing blocks.
Attachments
Attachments
Issue Links
- blocks
-
HDFS-10953 Remove single-rpc block reports.
- Open
- is broken by
-
HDFS-7960 The full block report should prune zombie storages even if they're not empty
- Closed