hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Matteo Bertozzi (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-16464) archive folder grows bigger and bigger due to corrupt snapshot under tmp dir
Date Mon, 22 Aug 2016 14:03:20 GMT

    [ https://issues.apache.org/jira/browse/HBASE-16464?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15430810#comment-15430810
] 

Matteo Bertozzi commented on HBASE-16464:
-----------------------------------------

SnapshotCleaner is a bad place to delete the .tmp dir, since the cleaner is not aware of "in-progress"
snapshots. you'll end up cancelling files from an in-progress snapshot and having the snapshot
request to fail.

we have a cleanup of the .tmp dir on snapshot failure.. why that does not work?
https://github.com/apache/hbase/blob/master/hbase-server/src/main/java/org/apache/hadoop/hbase/master/snapshot/TakeSnapshotHandler.java#L225

> archive folder grows bigger and bigger due to corrupt snapshot under tmp dir
> ----------------------------------------------------------------------------
>
>                 Key: HBASE-16464
>                 URL: https://issues.apache.org/jira/browse/HBASE-16464
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 1.1.1
>            Reporter: Heng Chen
>            Assignee: Heng Chen
>             Fix For: 2.0.0, 1.3.0, 1.4.0, 1.1.6, 1.2.3
>
>         Attachments: HBASE-16464-branch-1.1.patch, HBASE-16464.patch
>
>
> We met the problem on our real production cluster,  we need to cleanup some data on hbase,
 we notice the archive folder is much larger than others,  so we delete all snapshots of all
tables,  but the archive folder still grows bigger and bigger. 
> After check the hmaster log, we notice the exception below:
> {code}
> 2016-08-22 15:34:33,089 ERROR [f04,16000,1471240833208_ChoreService_1] snapshot.SnapshotHFileCleaner:
Exception while checking if files were valid, keeping them just in case.
> org.apache.hadoop.hbase.snapshot.CorruptedSnapshotException: Couldn't read snapshot info
from:hdfs://f04/hbase/.hbase-snapshot/.tmp/frog_stastic_2016-08-17/.snapshotinfo
>         at org.apache.hadoop.hbase.snapshot.SnapshotDescriptionUtils.readSnapshotInfo(SnapshotDescriptionUtils.java:295)
>         at org.apache.hadoop.hbase.snapshot.SnapshotReferenceUtil.getHFileNames(SnapshotReferenceUtil.java:328)
>         at org.apache.hadoop.hbase.master.snapshot.SnapshotHFileCleaner$1.filesUnderSnapshot(SnapshotHFileCleaner.java:85)
>         at org.apache.hadoop.hbase.master.snapshot.SnapshotFileCache.getSnapshotsInProgress(SnapshotFileCache.java:303)
>         at org.apache.hadoop.hbase.master.snapshot.SnapshotFileCache.getUnreferencedFiles(SnapshotFileCache.java:194)
>         at org.apache.hadoop.hbase.master.snapshot.SnapshotHFileCleaner.getDeletableFiles(SnapshotHFileCleaner.java:62)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteFiles(CleanerChore.java:233)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteEntries(CleanerChore.java:157)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteDirectory(CleanerChore.java:180)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteEntries(CleanerChore.java:149)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteDirectory(CleanerChore.java:180)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteEntries(CleanerChore.java:149)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteDirectory(CleanerChore.java:180)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteEntries(CleanerChore.java:149)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteDirectory(CleanerChore.java:180)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteEntries(CleanerChore.java:149)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteDirectory(CleanerChore.java:180)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.checkAndDeleteEntries(CleanerChore.java:149)
>         at org.apache.hadoop.hbase.master.cleaner.CleanerChore.chore(CleanerChore.java:124)
>         at org.apache.hadoop.hbase.ScheduledChore.run(ScheduledChore.java:185)
>         at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
>         at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
>         at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
>         at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
>         at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
>         at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
>         at java.lang.Thread.run(Thread.java:745)
> Caused by: java.io.FileNotFoundException: File does not exist: /hbase/.hbase-snapshot/.tmp/frog_stastic_2016-08-17/.snapshotinfo
>         at org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:71)
>         at org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:61)
>         at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsInt(FSNamesystem.java:1828)
>         at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocations(FSNamesystem.java:1799)
>         at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocations(FSNamesystem.java:1712)
>         at org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.getBlockLocations(NameNodeRpcServer.java:587)
>         at org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.getBlockLocations(ClientNamenodeProtocolServerSideTranslatorPB.java:365)
>         at org.apache.hadoop.hdfs.protocol.proto.ClientNamenodeProtocolProtos$ClientNamenodeProtocol$2.callBlockingMethod(ClientNamenodeProtocolProtos.java)
>         at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:616)
> {code}
> It means when SnapshotHFileCleaner begin to cleanup the archive folder, it reads the
the snapshot dir to check if any links to hfiles exist, but when read the file /.hbase-snapshot/.tmp/frog_stastic_2016-08-17/.snapshotinfo,
corrupt exception thrown out (not sure why the file not found),  and cleanup will be failed.
> When i check the /.hbase-snapshot/.tmp/frog_stastic_2016-08-17, i can see there is only
one file exist /hbase/.hbase-snapshot/.tmp/frog_stastic_2016-08-17/region-manifest.8e3179c388e10770eba7d35e30f2777f,
 /hbase/.hbase-snapshot/.tmp/frog_stastic_2016-08-17/.snapshotinfo missed. 
> I think we should catch up the exception and delete the file to ensure cleanup will go
on.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message