[HBASE-14885] NullPointerException in HMaster#normalizeRegions() due to missing TableDescriptor

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.3.0
• Fix Version/s: 1.2.0, 1.3.0, 2.0.0
• Component/s: master
• Hadoop Flags: Reviewed
• Assignee: Ted Yu
• Reporter: Ted Yu
• Created: 25/Nov/15 19:49
• Updated: 22/Dec/15 10:21
• Resolved: 26/Nov/15 00:11

Description

During system test on Windows, we observed the following in master log:

2015-11-23 11:31:38,853 ERROR org.apache.hadoop.hbase.master.normalizer.RegionNormalizerChore: Caught error^M
java.lang.NullPointerException^M
  at org.apache.hadoop.hbase.master.HMaster.normalizeRegions(HMaster.java:1396)^M
  at org.apache.hadoop.hbase.master.normalizer.RegionNormalizerChore.chore(RegionNormalizerChore.java:49)^M
  at org.apache.hadoop.hbase.ScheduledChore.run(ScheduledChore.java:185)^M
  at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)^M
  at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)^M
  at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)^M
  at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)^M
  at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)^M
  at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)^M
  at java.lang.Thread.run(Thread.java:745)^M

The NullPointerException came from the second line below:

      for(TableName table : allEnabledTables) {
        if (table.isSystemTable() || !getTableDescriptors().get(table).isNormalizationEnabled()) {

It seems TableDescriptor for some table was absent.

normalizeRegions() should deal with such scenario without producing NPE.