[HBASE-8295] TestMasterFailover.testMasterFailoverWithMockedRITOnDeadRS is flaky.

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.95.0
• Fix Version/s: 0.98.0, 0.95.1
• Hadoop Flags: Reviewed
• Assignee: rajeshbabu
• Reporter: rajeshbabu
• Created: 08/Apr/13 18:49
• Updated: 23/Sep/13 19:08
• Resolved: 23/Apr/13 19:22

Description

TestMasterFailover.testMasterFailoverWithMockedRITOnDeadRS failed because asserting before finishing SSH process of dead server.

https://builds.apache.org/job/PreCommit-HBASE-Build/5182//artifact/trunk/hbase-server/target/surefire-reports/org.apache.hadoop.hbase.master.TestMasterFailover.txt

In SSH Checked how many regions carrying by dead server but not started assignment before asserting.

2013-04-08 12:17:18,238 INFO  [MASTER_SERVER_OPERATIONS-asf001.sp2.ygridcore.net,41490,1365423432616-1] handler.ServerShutdownHandler(192): Reassigning 14 region(s) that asf001.sp2.ygridcore.net,46030,1365423422112 was carrying (and 0 regions(s) that were opening on this server)

We need to wait until SSH processing finished to ensure atleast regions assignment started. After that we will any way wait at blockUntilNoRIT until assignment completed.

Issue Links

• is related to: HBASE-8991 TestMasterFailover.testMasterFailoverWithMockedRITOnDeadRS failed again