[HIVE-1422] skip counter update when RunningJob.getCounters() returns null

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.6.0
• Fix Version/s: 0.7.0
• Component/s: Query Processor
• Hadoop Flags: Reviewed
• Assignee: Joydeep Sen Sarma
• Reporter: John Sichi
• Created: 22/Jun/10 01:33
• Updated: 17/Dec/11 00:00
• Resolved: 02/Aug/10 23:43

Description

Under heavy load circumstances on some Hadoop versions, we may get a NPE from trying to dereference a null Counters object. I don't have a unit test which can reproduce it, but here's an example stack from a production cluster we saw today:

10/06/21 13:01:10 ERROR exec.ExecDriver: Ended Job = job_201005200457_701060 with exception 'java.lang.NullPointerException(null)'
java.lang.NullPointerException
at org.apache.hadoop.hive.ql.exec.Operator.updateCounters(Operator.java:999)
at org.apache.hadoop.hive.ql.exec.ExecDriver.updateCounters(ExecDriver.java:503)
at org.apache.hadoop.hive.ql.exec.ExecDriver.progress(ExecDriver.java:390)
at org.apache.hadoop.hive.ql.exec.ExecDriver.execute(ExecDriver.java:697)
at org.apache.hadoop.hive.ql.exec.Task.executeTask(Task.java:107)
at org.apache.hadoop.hive.ql.exec.TaskRunner.runSequential(TaskRunner.java:55)
at org.apache.hadoop.hive.ql.exec.TaskRunner.run(TaskRunner.java:47)

Issue Links

• incorporates: HIVE-1493 incorrect explanation when local mode not chosen automatically