hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "rajeshbabu (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-8814) Possible NPE in split if a region has empty store files.
Date Fri, 28 Jun 2013 18:44:20 GMT

    [ https://issues.apache.org/jira/browse/HBASE-8814?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13695655#comment-13695655
] 

rajeshbabu commented on HBASE-8814:
-----------------------------------

bq. The store file(which u refer) is a result of a major compaction where all the puts are
removed by later deletes right?  Does the test case ensures this?
Yes Anoop. In test case I didn't trigger major compaction from admin but its happening because
all store files(there are less than the configured maximum files) are eligible for compaction.
                
> 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
>
>         Attachments: HBASE-8814_trunk.patch
>
>
> {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

Mime
View raw message