Return-Path: X-Original-To: apmail-hadoop-hdfs-issues-archive@minotaur.apache.org Delivered-To: apmail-hadoop-hdfs-issues-archive@minotaur.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id CE92B18D7D for ; Thu, 3 Dec 2015 02:10:11 +0000 (UTC) Received: (qmail 42743 invoked by uid 500); 3 Dec 2015 02:10:11 -0000 Delivered-To: apmail-hadoop-hdfs-issues-archive@hadoop.apache.org Received: (qmail 42685 invoked by uid 500); 3 Dec 2015 02:10:11 -0000 Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: hdfs-issues@hadoop.apache.org Delivered-To: mailing list hdfs-issues@hadoop.apache.org Received: (qmail 42436 invoked by uid 99); 3 Dec 2015 02:10:11 -0000 Received: from arcas.apache.org (HELO arcas) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 03 Dec 2015 02:10:11 +0000 Received: from arcas.apache.org (localhost [127.0.0.1]) by arcas (Postfix) with ESMTP id 472B62C1F61 for ; Thu, 3 Dec 2015 02:10:11 +0000 (UTC) Date: Thu, 3 Dec 2015 02:10:11 +0000 (UTC) From: "Tsz Wo Nicholas Sze (JIRA)" To: hdfs-issues@hadoop.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Updated] (HDFS-9294) DFSClient deadlock when close file and failed to renew lease MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394 [ https://issues.apache.org/jira/browse/HDFS-9294?page=3Dcom.atlassian= .jira.plugin.system.issuetabpanels:all-tabpanel ] Tsz Wo Nicholas Sze updated HDFS-9294: -------------------------------------- Resolution: Fixed Fix Version/s: 2.7.2 Status: Resolved (was: Patch Available) I have committed this. Thanks, Brahma! > DFSClient deadlock when close file and failed to renew lease > ------------------------------------------------------------- > > Key: HDFS-9294 > URL: https://issues.apache.org/jira/browse/HDFS-9294 > Project: Hadoop HDFS > Issue Type: Bug > Components: hdfs-client > Affects Versions: 2.2.0, 2.7.1 > Environment: Hadoop 2.2.0 > Reporter: =E9=82=93=E9=A3=9E > Assignee: Brahma Reddy Battula > Priority: Blocker > Fix For: 2.7.2 > > Attachments: HDFS-9294-002.patch, HDFS-9294-002.patch, HDFS-9294-= branch-2.7.patch, HDFS-9294-branch-2.patch, HDFS-9294.patch > > > We found a deadlock at our HBase(0.98) cluster(and the Hadoop Version is = 2.2.0),and it should be HDFS BUG,at the time our network is not stable. > below is the stack: > *************************************************************************= ************************************************************ > Found one Java-level deadlock: > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > "MemStoreFlusher.1": > waiting to lock monitor 0x00007ff27cfa5218 (object 0x00000002fae5ebe0, = a org.apache.hadoop.hdfs.LeaseRenewer), > which is held by "LeaseRenewer:hbaseadmin@hbase-ns-gdt-sh-marvel" > "LeaseRenewer:hbaseadmin@hbase-ns-gdt-sh-marvel": > waiting to lock monitor 0x00007ff2e67e16a8 (object 0x0000000486ce6620, = a org.apache.hadoop.hdfs.DFSOutputStream), > which is held by "MemStoreFlusher.0" > "MemStoreFlusher.0": > waiting to lock monitor 0x00007ff27cfa5218 (object 0x00000002fae5ebe0, = a org.apache.hadoop.hdfs.LeaseRenewer), > which is held by "LeaseRenewer:hbaseadmin@hbase-ns-gdt-sh-marvel" > Java stack information for the threads listed above: > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D > "MemStoreFlusher.1": > =09at org.apache.hadoop.hdfs.LeaseRenewer.addClient(LeaseRenewer.java:216= ) > =09- waiting to lock <0x00000002fae5ebe0> (a org.apache.hadoop.hdfs.Lease= Renewer) > =09at org.apache.hadoop.hdfs.LeaseRenewer.getInstance(LeaseRenewer.java:8= 1) > =09at org.apache.hadoop.hdfs.DFSClient.getLeaseRenewer(DFSClient.java:648= ) > =09at org.apache.hadoop.hdfs.DFSClient.endFileLease(DFSClient.java:659) > =09at org.apache.hadoop.hdfs.DFSOutputStream.close(DFSOutputStream.java:1= 882) > =09- locked <0x000000055b606cb0> (a org.apache.hadoop.hdfs.DFSOutputStrea= m) > =09at org.apache.hadoop.fs.FSDataOutputStream$PositionCache.close(FSDataO= utputStream.java:71) > =09at org.apache.hadoop.fs.FSDataOutputStream.close(FSDataOutputStream.ja= va:104) > =09at org.apache.hadoop.hbase.io.hfile.AbstractHFileWriter.finishClose(Ab= stractHFileWriter.java:250) > =09at org.apache.hadoop.hbase.io.hfile.HFileWriterV2.close(HFileWriterV2.= java:402) > =09at org.apache.hadoop.hbase.regionserver.StoreFile$Writer.close(StoreFi= le.java:974) > =09at org.apache.hadoop.hbase.regionserver.StoreFlusher.finalizeWriter(St= oreFlusher.java:78) > =09at org.apache.hadoop.hbase.regionserver.DefaultStoreFlusher.flushSnaps= hot(DefaultStoreFlusher.java:75) > =09- locked <0x000000059869eed8> (a java.lang.Object) > =09at org.apache.hadoop.hbase.regionserver.HStore.flushCache(HStore.java:= 812) > =09at org.apache.hadoop.hbase.regionserver.HStore$StoreFlusherImpl.flushC= ache(HStore.java:1974) > =09at org.apache.hadoop.hbase.regionserver.HRegion.internalFlushcache(HRe= gion.java:1795) > =09at org.apache.hadoop.hbase.regionserver.HRegion.internalFlushcache(HRe= gion.java:1678) > =09at org.apache.hadoop.hbase.regionserver.HRegion.flushcache(HRegion.jav= a:1591) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher.flushRegion(Me= mStoreFlusher.java:472) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher.flushOneForGlo= balPressure(MemStoreFlusher.java:211) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher.access$500(Mem= StoreFlusher.java:66) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher$FlushHandler.r= un(MemStoreFlusher.java:238) > =09at java.lang.Thread.run(Thread.java:744) > "LeaseRenewer:hbaseadmin@hbase-ns-gdt-sh-marvel": > =09at org.apache.hadoop.hdfs.DFSOutputStream.abort(DFSOutputStream.java:1= 822) > =09- waiting to lock <0x0000000486ce6620> (a org.apache.hadoop.hdfs.DFSOu= tputStream) > =09at org.apache.hadoop.hdfs.DFSClient.closeAllFilesBeingWritten(DFSClien= t.java:780) > =09at org.apache.hadoop.hdfs.DFSClient.abort(DFSClient.java:753) > =09at org.apache.hadoop.hdfs.LeaseRenewer.run(LeaseRenewer.java:453) > =09- locked <0x00000002fae5ebe0> (a org.apache.hadoop.hdfs.LeaseRenewer) > =09at org.apache.hadoop.hdfs.LeaseRenewer.access$700(LeaseRenewer.java:71= ) > =09at org.apache.hadoop.hdfs.LeaseRenewer$1.run(LeaseRenewer.java:298) > =09at java.lang.Thread.run(Thread.java:744) > "MemStoreFlusher.0": > =09at org.apache.hadoop.hdfs.LeaseRenewer.addClient(LeaseRenewer.java:216= ) > =09- waiting to lock <0x00000002fae5ebe0> (a org.apache.hadoop.hdfs.Lease= Renewer) > =09at org.apache.hadoop.hdfs.LeaseRenewer.getInstance(LeaseRenewer.java:8= 1) > =09at org.apache.hadoop.hdfs.DFSClient.getLeaseRenewer(DFSClient.java:648= ) > =09at org.apache.hadoop.hdfs.DFSClient.endFileLease(DFSClient.java:659) > =09at org.apache.hadoop.hdfs.DFSOutputStream.close(DFSOutputStream.java:1= 882) > =09- locked <0x0000000486ce6620> (a org.apache.hadoop.hdfs.DFSOutputStrea= m) > =09at org.apache.hadoop.fs.FSDataOutputStream$PositionCache.close(FSDataO= utputStream.java:71) > =09at org.apache.hadoop.fs.FSDataOutputStream.close(FSDataOutputStream.ja= va:104) > =09at org.apache.hadoop.hbase.io.hfile.AbstractHFileWriter.finishClose(Ab= stractHFileWriter.java:250) > =09at org.apache.hadoop.hbase.io.hfile.HFileWriterV2.close(HFileWriterV2.= java:402) > =09at org.apache.hadoop.hbase.regionserver.StoreFile$Writer.close(StoreFi= le.java:974) > =09at org.apache.hadoop.hbase.regionserver.StoreFlusher.finalizeWriter(St= oreFlusher.java:78) > =09at org.apache.hadoop.hbase.regionserver.DefaultStoreFlusher.flushSnaps= hot(DefaultStoreFlusher.java:75) > =09- locked <0x00000004888f6848> (a java.lang.Object) > =09at org.apache.hadoop.hbase.regionserver.HStore.flushCache(HStore.java:= 812) > =09at org.apache.hadoop.hbase.regionserver.HStore$StoreFlusherImpl.flushC= ache(HStore.java:1974) > =09at org.apache.hadoop.hbase.regionserver.HRegion.internalFlushcache(HRe= gion.java:1795) > =09at org.apache.hadoop.hbase.regionserver.HRegion.internalFlushcache(HRe= gion.java:1678) > =09at org.apache.hadoop.hbase.regionserver.HRegion.flushcache(HRegion.jav= a:1591) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher.flushRegion(Me= mStoreFlusher.java:472) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher.flushRegion(Me= mStoreFlusher.java:435) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher.access$800(Mem= StoreFlusher.java:66) > =09at org.apache.hadoop.hbase.regionserver.MemStoreFlusher$FlushHandler.r= un(MemStoreFlusher.java:253) > =09at java.lang.Thread.run(Thread.java:744) > Found 1 deadlock.=20 > ********************************************************************** > the thread "MemStoreFlusher.0" is closing outputStream and remove it's le= ase ; > other side the daemon thread "LeaseRenewer" failed to connect active nn = for renewing lease,but got SocketTimeoutException cause of network is n= ot good,so abort outputstream. > then deadlock is made. > and it seems not solved at Hadoop 2.7.1 .If confirmed , we can fixed the = issue. -- This message was sent by Atlassian JIRA (v6.3.4#6332)