[HBASE-19065] HRegion#bulkLoadHFiles() should wait for concurrent Region#flush() to finish

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.4.0, 1.3.2, 2.0.0-beta-1, 2.0.0, 1.2.7
• Hadoop Flags: Reviewed
• Assignee: Ted Yu
• Reporter: Ted Yu
• Created: 22/Oct/17 17:26
• Updated: 01/Aug/18 06:24
• Resolved: 01/Nov/17 16:26

Description

When I was debugging bulk load failure, I saw the following in region server log:

2017-10-17 23:05:28,795 DEBUG [B.defaultRpcServer.handler=0,queue=0,port=16020] regionserver.HRegion: NOT flushing memstore for region mx_,         f449669a8b0341e4edbd2ebdacc72094f449669a8b0341e4edbd2ebdacc7209420150711,1504909319142.52d496ba39036e0c2cc9522895ad438f., flushing=true, writesEnabled=true
2017-10-17 23:05:28,796 ERROR [B.defaultRpcServer.handler=0,queue=0,port=16020] access.SecureBulkLoadEndpoint: Failed to complete bulk load
java.io.IOException: Could not bulk load with an assigned sequential ID because the flush didn't run. Reason for not flushing: Not flushing since already flushing
  at org.apache.hadoop.hbase.regionserver.HRegion.bulkLoadHFiles(HRegion.java:5282)
  at org.apache.hadoop.hbase.security.access.SecureBulkLoadEndpoint$1.run(SecureBulkLoadEndpoint.java:292)
  at org.apache.hadoop.hbase.security.access.SecureBulkLoadEndpoint$1.run(SecureBulkLoadEndpoint.java:275)

There was concurrent flush which got misinterpreted by bulkLoadHFiles().
HRegion#bulkLoadHFiles() should wait for the concurrent flush to complete.

Issue Links

• relates to: HBASE-19442 Backport HBASE-19065 to branch-1 (HRegion#bulkLoadHFiles() should wait for concurrent Region#flush() to finish)