[HADOOP-8220] ZKFailoverController doesn't handle failure to become active correctly

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 0.23.3, 0.24.0
• Fix Version/s: Auto Failover (HDFS-3042)
• Component/s: auto-failover, ha
• Target Version/s: Auto Failover (HDFS-3042)
• Hadoop Flags: Reviewed
• Assignee: Todd Lipcon
• Reporter: Todd Lipcon
• Created: 27/Mar/12 04:29
• Updated: 30/Mar/12 20:24
• Resolved: 30/Mar/12 20:24

Description

The ZKFC doesn't properly handle the case where the monitored service fails to become active. Currently, it catches the exception and logs a warning, but then continues on, after calling quitElection(). This causes a NPE when it later tries to use the same zkClient instance while handling that same request. There is a test case, but the test case doesn't ensure that the node that had the failure is later able to recover properly.