[HBASE-3278] AssertionError in LoadBalancer

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.90.0, 0.92.0
• Hadoop Flags: Reviewed
• Assignee: Jonathan Gray
• Reporter: Jean-Daniel Cryans
• Created: 24/Nov/10 23:53
• Updated: 20/Nov/15 12:43
• Resolved: 07/Dec/10 05:27

Description

While running PE with low splitting configuration, I got this:

2010-11-24 23:23:24,508 INFO org.apache.hadoop.hbase.master.ServerManager: Received REGION_SPLIT: TestTable,0002485653,...
2010-11-24 23:23:26,129 INFO org.apache.hadoop.hbase.master.ServerManager: Received REGION_SPLIT: TestTable,0004309306,...
2010-11-24 23:23:26,132 INFO org.apache.hadoop.hbase.master.ServerManager: Received REGION_SPLIT: TestTable,0001281491...
2010-11-24 23:23:26,162 FATAL org.apache.hadoop.hbase.master.HMaster$1: sv2borg180:61000-BalancerChoreerror
java.lang.AssertionError
	at org.apache.hadoop.hbase.master.LoadBalancer.balanceCluster(LoadBalancer.java:296)
	at org.apache.hadoop.hbase.master.HMaster.balance(HMaster.java:679)
	at org.apache.hadoop.hbase.master.HMaster$1.chore(HMaster.java:578)
	at org.apache.hadoop.hbase.Chore.run(Chore.java:66)
2010-11-24 23:23:26,163 INFO org.apache.hadoop.hbase.master.HMaster$1: sv2borg180:61000-BalancerChore exiting
2010-11-24 23:23:26,236 INFO org.apache.hadoop.hbase.master.ServerManager: Received REGION_SPLIT: TestTable,0000921369,...

The thread dies but the master survives. There's nothing specific before that in the log, just regions splitting.

The line in LoadBalancer is:

    assert(regionidx == regionsToMove.size());