You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Arwin S Tio (JIRA)" <ji...@apache.org> on 2019/07/23 06:26:00 UTC
[jira] [Commented] (HBASE-16464) archive folder grows bigger and
bigger due to corrupt snapshot under tmp dir
[ https://issues.apache.org/jira/browse/HBASE-16464?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16890700#comment-16890700 ]
Arwin S Tio commented on HBASE-16464:
-------------------------------------
Hello all,
We are running on an ancient version of HBase unfortunately and have run into this issue aswell.
[~mbertozzi] You mentioned:
{quote}my guess is that the snapshot may have timed out. TakeSnapshotHandler had cleaned .tmp but one RS (the one with 8e3179c388e10770eba7d35e30f2777f) was still doing trying to snapshot after timeout. HRegion#addRegionToSnapshot() writes the region-manifest you have in .tmp. and that will never be cleaned since TakeSnapshotHandler is already done.
{quote}
May I ask if the snapshot timeouts imply that there is data loss/inconsistency? i.e. if the RS completed the Snapshot after TakeSnapshotHandler cleaned up, is the snapshot data for that RS missing/inconsistent?
Thanks
> 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
> Components: snapshots
> Affects Versions: 1.1.1
> Reporter: Heng Chen
> Assignee: Heng Chen
> Priority: Major
> Labels: reviewed
> Fix For: 1.3.0, 1.1.6, 1.2.3, 0.98.22, 2.0.0
>
> Attachments: HBASE-16464-branch-1.1.patch, HBASE-16464.patch, HBASE-16464.v1.patch, HBASE-16464.v1.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
(v7.6.14#76016)