Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Thu, 27 Jun 2013 08:58:20 +0000 (UTC)
From: "rajeshbabu (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <JIRA.12655097.1372323441770.179000.1372323500034@arcas>
In-Reply-To: <JIRA.12655097.1372323441770@arcas>
References: <JIRA.12655097.1372323441770@arcas>
Subject: [jira] [Created] (HBASE-8814) Possible NPE in split if a region has
 empty store files.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

rajeshbabu created HBASE-8814:
---------------------------------

             Summary: Possible NPE in split if a region has empty store files.
                 Key: HBASE-8814
                 URL: https://issues.apache.org/jira/browse/HBASE-8814
             Project: HBase
          Issue Type: Bug
          Components: regionserver
            Reporter: rajeshbabu
            Assignee: rajeshbabu
             Fix For: 0.98.0, 0.95.2, 0.94.10


{code}
2013-06-27 14:12:54,472 INFO  [RS:1;BLRY2R009039160:49833-splits-1372322572806] regionserver.SplitRequest(92): Running rollback/cleanup of failed split of testSplitShouldNotThrowNPEEvenARegionHasEmptySplitFiles,,1372322556662.276e00da1420119e2f91f3a4c4c41d78.; java.util.concurrent.ExecutionException: java.lang.NullPointerException
java.io.IOException: java.util.concurrent.ExecutionException: java.lang.NullPointerException
	at org.apache.hadoop.hbase.regionserver.SplitTransaction.splitStoreFiles(SplitTransaction.java:602)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction.createDaughters(SplitTransaction.java:297)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction.execute(SplitTransaction.java:466)
	at org.apache.hadoop.hbase.regionserver.SplitRequest.run(SplitRequest.java:82)
	at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
	at java.lang.Thread.run(Thread.java:662)
Caused by: java.util.concurrent.ExecutionException: java.lang.NullPointerException
	at java.util.concurrent.FutureTask$Sync.innerGet(FutureTask.java:222)
	at java.util.concurrent.FutureTask.get(FutureTask.java:83)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction.splitStoreFiles(SplitTransaction.java:596)
	... 6 more
Caused by: java.lang.NullPointerException
	at org.apache.hadoop.hbase.regionserver.HRegionFileSystem.splitStoreFile(HRegionFileSystem.java:539)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction.splitStoreFile(SplitTransaction.java:610)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction.access$1(SplitTransaction.java:607)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction$StoreFileSplitter.call(SplitTransaction.java:633)
	at org.apache.hadoop.hbase.regionserver.SplitTransaction$StoreFileSplitter.call(SplitTransaction.java:1)
	at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
	at java.util.concurrent.FutureTask.run(FutureTask.java:138)
	... 3 more
{code}

If a storefile is empty(can be because of puts and deletes) then first and lastkey of the file will be empty. Then we will get first or last key as null. Then we will end up in NPE when we will check splitkey in the range or not.
{code}
    if (top) {
      //check if larger than last key.
      KeyValue splitKey = KeyValue.createFirstOnRow(splitRow);
      byte[] lastKey = f.createReader().getLastKey();      
      if (f.getReader().getComparator().compare(splitKey.getBuffer(), 
          splitKey.getKeyOffset(), splitKey.getKeyLength(), lastKey, 0, lastKey.length) > 0) {
        return null;
      }
    } else {
      //check if smaller than first key
      KeyValue splitKey = KeyValue.createLastOnRow(splitRow);
      byte[] firstKey = f.createReader().getFirstKey();
      if (f.getReader().getComparator().compare(splitKey.getBuffer(), 
          splitKey.getKeyOffset(), splitKey.getKeyLength(), firstKey, 0, firstKey.length) < 0) {
        return null;
      }      
    }
{code}

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira