Hi Kihwal, Thank you for the response. This led to us deep diving into the system level I/O behavior (with help from some resident ops experts), and we definitely identified some bottlenecks. We also realized we had recently increased the RAM on the NN servers, so the allowable dirty page cache size would've increased proportionally.
We've been working on the disk partition setup and tuning the dirty page ratio settings and have been seeing some improvements. Definitely seems like we're on the right track now. We'll be pushing hard to get SSDs in these as well. Apologies for the false alarm/distraction. Thank you again for your help! Joey On Tue, Sep 13, 2016 at 11:12 AM, Kihwal Lee <kih...@yahoo-inc.com> wrote: > Is the system busy with I/O when it happens? Any other I/O activities > preceding the event? In your case DistCp could have generated extra edits > and also namenode daemon and audit log entries. Depending on > configuration, dirty pages can pile up quite a bit on Linux systems with a > large memory and cause extreme I/O delays when they hit the drive. fsimage > uploading might be contributing to that. But we haven't seen any issues > like that. In one of large clusters (5000+ node, 2.7.3ish, jdk8), > rollEdits() takes less than 30ms consistently. > > Kihwal > > > ------------------------------ > *From:* Joey Paskhay <joey.pask...@gmail.com> > *To:* hdfs-dev@hadoop.apache.org > *Sent:* Tuesday, September 13, 2016 12:06 PM > *Subject:* Active NameNode Image Download and Edit Log Roll Stuck in > FileChannel Force/Truncate > > Reposting here to see if any of the HDFS developers have some good insight > into this. > > Deep dive is in the below original message. The gist of it is after > upgrading to 2.7.2 on a ~260 node cluster, the active NN's fsimage download > and edit logs roll seem to get stuck in native FileChannel.force calls > (sometimes FileChannel.truncate). This leads to the ZKFC health monitor > failing (all the RPC handler threads back up waiting for the > FSNamesystem.fsLock to be released by the edit log roll process), and the > active NN gets killed. > > Happens occasionally when the system is idle (once a day) but very > frequently when we run DistCp (every 20-30 minutes). We believe we saw this > every month or two on 2.2.1 (logs/files rolled over since last time so > can't confirm exact same issue), but with 2.7.2 it seems to be much more > frequent. > > Any help or guidance would be much appreciated. > > Thanks, > Joey > > > Hey there, > > We're in the process of upgrading our Hadoop cluster from 2.2.1 to 2.7.2 > and currently testing 2.7.2 in our pre-prod/backup cluster. We're seeing a > lot of active NameNode failovers (sometimes as often as every 30 minutes), > especially when we're running DistCp to copy data from our production > cluster for users to test with. We had seen similar failovers occasionally > while running 2.2.1, but not nearly as often (once every month or two). > Haven't been able to verify it's the exact same root cause in the 2.2.1 > version since files/logs have rolled over since the last time it happened. > > So here's the chain of events we've found so far. Hoping someone can > provide further direction. > > The standby NameNode's checkpointing process succeeds locally and issues > the image PUT request in TransferFsImage.uploadImage. The active NameNode > finishes downloading the fsimage.ckpt file, but when it tries to issue > the fos.getChannel().force(true) call in TransferFsImage.receiveFile it > seems to get stuck in native code. The standby NameNode then gets a > SocketTimeoutException -- it happens 60 seconds after the last modification > time we see in the "stat" output for the fsimage.ckpt file that the active > NameNode pulled down. > > Right after the time this is happening (~30 sec after the last modification > to the fsimage.ckpt file) we see a similar issue with the edit log roll. > The standby NameNode's EditLogTailer triggers the rolling of the edit log > on the active NameNode. We see the active NameNode enter its rollEditLog > process, and will either see the endCurrentLogSegment call get stuck in > EditLogFileOutputStream.close on the fc.truncate(fc.position()) call or the > startLogSegment call get stuck in EditLogFileOutputStream.flushAndSync on > the fc.force(true) call. They both get stuck in the native code. Looking at > the last modification time in the "stat" output of the edits file, we see > that 20 seconds later the standby NameNode's RPC call times out. > > The rollEditLog ends up holding onto the FSNamesystem's write lock on > fsLock, and this causes all other RPC calls to pile up trying to acquire > read locks until ZKFC times out on the health monitor and signals for the > NameNode to be killed. We patched the SshFenceByTcpPort code to issue a > kill -3 to get a thread dump before it kills the active NameNode. > > We're running on CentOS 6 using ext4 FS (w/ noatime) using kernel 2.6.32. > The fsimage file is typically ~7.2GB and the edits files are typically > ~1MB-2MB. The cluster running 2.7.2 is 256 nodes. We're running on > JDK 1.8.0_92 (compiled against it too w/ a few JDK8 specific patches). > > See the relevant stacks below of the FileChannel code getting stuck in the > native code. I can also provide the full thread dumps and any relevant > configs, if needed. > > Tried looking in JIRA and online but didn't see anything directly related. > Any insight as to whether this is a bug in Hadoop or if it's a side-effect > of something else? When the cluster is mostly idle, everything seems fine. > Our dev/test clusters haven't had any issues with the upgrades but they're > only 10 nodes or less and have little load. > > Thanks! > Joey > > Example of both getting stuck in force calls: > > "641242166@qtp-1147805316-11" #869 daemon prio=5 os_prio=0 > tid=0x00007fd9c3bc4800 nid=0x2f37 runnable [0x00007fb7f6c8d000] > java.lang.Thread.State: RUNNABLE > at sun.nio.ch.FileDispatcherImpl.force0(Native Method) > at sun.nio.ch.FileDispatcherImpl.force(FileDispatcherImpl.java:76) > at sun.nio.ch.FileChannelImpl.force(FileChannelImpl.java:388) > at org.apache.hadoop.hdfs.server.namenode.TransferFsImage. > receiveFile(TransferFsImage.java:530) > at org.apache.hadoop.hdfs.server.namenode.TransferFsImage. > handleUploadImageRequest(TransferFsImage.java:132) > at org.apache.hadoop.hdfs.server.namenode.ImageServlet$2.run( > ImageServlet.java:488) > at org.apache.hadoop.hdfs.server.namenode.ImageServlet$2.run( > ImageServlet.java:458) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:422) > at org.apache.hadoop.security.UserGroupInformation.doAs( > UserGroupInformation.java:1657) > at org.apache.hadoop.hdfs.server.namenode.ImageServlet.doPut( > ImageServlet.java:457) > at javax.servlet.http.HttpServlet.service(HttpServlet.java:730) > at javax.servlet.http.HttpServlet.service(HttpServlet.java:820) > at org.mortbay.jetty.servlet.ServletHolder.handle( > ServletHolder.java:511) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1221) > at org.apache.hadoop.security.authentication.server. > AuthenticationFilter.doFilter(AuthenticationFilter.java:595) > at org.apache.hadoop.security.authentication.server. > AuthenticationFilter.doFilter(AuthenticationFilter.java:554) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.apache.hadoop.http.HttpServer2$QuotingInputFilter.doFilter( > HttpServer2.java:1243) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.apache.hadoop.http.NoCacheFilter.doFilter( > NoCacheFilter.java:45) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.apache.hadoop.http.NoCacheFilter.doFilter( > NoCacheFilter.java:45) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.mortbay.jetty.servlet.ServletHandler.handle( > ServletHandler.java:399) > at org.mortbay.jetty.security.SecurityHandler.handle( > SecurityHandler.java:216) > at org.mortbay.jetty.servlet.SessionHandler.handle( > SessionHandler.java:182) > at org.mortbay.jetty.handler.ContextHandler.handle( > ContextHandler.java:766) > at org.mortbay.jetty.webapp.WebAppContext.handle( > WebAppContext.java:450) > at org.mortbay.jetty.handler.ContextHandlerCollection.handle( > ContextHandlerCollection.java:230) > at org.mortbay.jetty.handler.HandlerWrapper.handle( > HandlerWrapper.java:152) > at org.mortbay.jetty.Server.handle(Server.java:326) > at org.mortbay.jetty.HttpConnection.handleRequest( > HttpConnection.java:542) > at org.mortbay.jetty.HttpConnection$RequestHandler. > content(HttpConnection.java:945) > at org.mortbay.jetty.HttpParser.parseNext(HttpParser.java:843) > at org.mortbay.jetty.HttpParser.parseAvailable(HttpParser. > java:212) > at org.mortbay.jetty.HttpConnection.handle( > HttpConnection.java:404) > at org.mortbay.io.nio.SelectChannelEndPoint.run( > SelectChannelEndPoint.java:410) > at org.mortbay.thread.QueuedThreadPool$PoolThread. > run(QueuedThreadPool.java:582) > > > "IPC Server handler 111 on 8020" #166 daemon prio=5 os_prio=0 > tid=0x00007fd9c3449000 nid=0x600d runnable [0x00007fb7fb1af000] > java.lang.Thread.State: RUNNABLE > at sun.nio.ch.FileDispatcherImpl.force0(Native Method) > at sun.nio.ch.FileDispatcherImpl.force(FileDispatcherImpl.java:76) > at sun.nio.ch.FileChannelImpl.force(FileChannelImpl.java:388) > at org.apache.hadoop.hdfs.server.namenode.EditLogFileOutputStream. > flushAndSync(EditLogFileOutputStream.java:206) > at org.apache.hadoop.hdfs.server.namenode.EditLogOutputStream. > flush(EditLogOutputStream.java:113) > at org.apache.hadoop.hdfs.server.namenode.EditLogOutputStream. > flush(EditLogOutputStream.java:107) > at org.apache.hadoop.hdfs.server.namenode.EditLogFileOutputStream. > create(EditLogFileOutputStream.java:122) > at org.apache.hadoop.hdfs.server.namenode.FileJournalManager. > startLogSegment(FileJournalManager.java:115) > - locked <0x00007fc671be42b0> (a org.apache.hadoop.hdfs.server. > namenode.FileJournalManager) > at org.apache.hadoop.hdfs.server.namenode.JournalSet$ > JournalAndStream.startLogSegment(JournalSet.java:107) > at org.apache.hadoop.hdfs.server.namenode.JournalSet$3.apply( > JournalSet.java:222) > at org.apache.hadoop.hdfs.server.namenode.JournalSet. > mapJournalsAndReportErrors(JournalSet.java:393) > at org.apache.hadoop.hdfs.server.namenode.JournalSet. > startLogSegment(JournalSet.java:219) > at org.apache.hadoop.hdfs.server.namenode.FSEditLog. > startLogSegment(FSEditLog.java:1237) > - locked <0x00007fba94879c00> (a org.apache.hadoop.hdfs.server. > namenode.FSEditLog) > at org.apache.hadoop.hdfs.server.namenode.FSEditLog. > rollEditLog(FSEditLog.java:1206) > - locked <0x00007fba94879c00> (a org.apache.hadoop.hdfs.server. > namenode.FSEditLog) > at org.apache.hadoop.hdfs.server.namenode.FSImage.rollEditLog( > FSImage.java:1294) > at org.apache.hadoop.hdfs.server.namenode.FSNamesystem. > rollEditLog(FSNamesystem.java:5832) > at org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer. > rollEditLog(NameNodeRpcServer.java:1121) > at org.apache.hadoop.hdfs.protocolPB. > NamenodeProtocolServerSideTran > slatorPB.rollEditLog(NamenodeProtocolServerSideTranslatorPB.java:142) > at org.apache.hadoop.hdfs.protocol.proto.NamenodeProtocolProtos$ > NamenodeProtocolService$2.callBlockingMethod(NamenodeProtocolProtos.java: > 12025) > at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ > ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:616) > at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:969) > at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2049) > at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2045) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:422) > at org.apache.hadoop.security.UserGroupInformation.doAs( > UserGroupInformation.java:1657) > at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2045) > > > Examples of rollEditLog getting stuck in truncate call and TransferFsImage > getting stuck in force call: > > "281983328@qtp-1147805316-22" #1085 daemon prio=5 os_prio=0 > tid=0x0000000061637000 nid=0x58a4 runnable [0x00007f5da10cd000] > java.lang.Thread.State: RUNNABLE > at sun.nio.ch.FileDispatcherImpl.force0(Native Method) > at sun.nio.ch.FileDispatcherImpl.force(FileDispatcherImpl.java:76) > at sun.nio.ch.FileChannelImpl.force(FileChannelImpl.java:388) > at org.apache.hadoop.hdfs.server.namenode.TransferFsImage. > receiveFile(TransferFsImage.java:530) > at org.apache.hadoop.hdfs.server.namenode.TransferFsImage. > handleUploadImageRequest(TransferFsImage.java:132) > at org.apache.hadoop.hdfs.server.namenode.ImageServlet$2.run( > ImageServlet.java:488) > at org.apache.hadoop.hdfs.server.namenode.ImageServlet$2.run( > ImageServlet.java:458) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:422) > at org.apache.hadoop.security.UserGroupInformation.doAs( > UserGroupInformation.java:1657) > at org.apache.hadoop.hdfs.server.namenode.ImageServlet.doPut( > ImageServlet.java:457) > at javax.servlet.http.HttpServlet.service(HttpServlet.java:730) > at javax.servlet.http.HttpServlet.service(HttpServlet.java:820) > at org.mortbay.jetty.servlet.ServletHolder.handle( > ServletHolder.java:511) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1221) > at org.apache.hadoop.security.authentication.server. > AuthenticationFilter.doFilter(AuthenticationFilter.java:595) > at org.apache.hadoop.security.authentication.server. > AuthenticationFilter.doFilter(AuthenticationFilter.java:554) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.apache.hadoop.http.HttpServer2$QuotingInputFilter.doFilter( > HttpServer2.java:1243) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.apache.hadoop.http.NoCacheFilter.doFilter( > NoCacheFilter.java:45) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.apache.hadoop.http.NoCacheFilter.doFilter( > NoCacheFilter.java:45) > at org.mortbay.jetty.servlet.ServletHandler$CachedChain. > doFilter(ServletHandler.java:1212) > at org.mortbay.jetty.servlet.ServletHandler.handle( > ServletHandler.java:399) > at org.mortbay.jetty.security.SecurityHandler.handle( > SecurityHandler.java:216) > at org.mortbay.jetty.servlet.SessionHandler.handle( > SessionHandler.java:182) > at org.mortbay.jetty.handler.ContextHandler.handle( > ContextHandler.java:766) > at org.mortbay.jetty.webapp.WebAppContext.handle( > WebAppContext.java:450) > at org.mortbay.jetty.handler.ContextHandlerCollection.handle( > ContextHandlerCollection.java:230) > at org.mortbay.jetty.handler.HandlerWrapper.handle( > HandlerWrapper.java:152) > at org.mortbay.jetty.Server.handle(Server.java:326) > at org.mortbay.jetty.HttpConnection.handleRequest( > HttpConnection.java:542) > at org.mortbay.jetty.HttpConnection$RequestHandler. > content(HttpConnection.java:945) > at org.mortbay.jetty.HttpParser.parseNext(HttpParser.java:843) > at org.mortbay.jetty.HttpParser.parseAvailable(HttpParser. > java:218) > at org.mortbay.jetty.HttpConnection.handle( > HttpConnection.java:404) > at org.mortbay.io.nio.SelectChannelEndPoint.run( > SelectChannelEndPoint.java:410) > at org.mortbay.thread.QueuedThreadPool$PoolThread. > run(QueuedThreadPool.java:582) > > > "IPC Server handler 41 on 8020" #97 daemon prio=5 os_prio=0 > tid=0x00007f5dbbdc6800 nid=0x611 runnable [0x00007f5de0ad0000] > java.lang.Thread.State: RUNNABLE > at sun.nio.ch.FileDispatcherImpl.truncate0(Native Method) > at sun.nio.ch.FileDispatcherImpl.truncate(FileDispatcherImpl. > java:80) > at sun.nio.ch.FileChannelImpl.truncate(FileChannelImpl.java:357) > - locked <0x00007f7bf1c4e220> (a java.lang.Object) > at org.apache.hadoop.hdfs.server.namenode. > EditLogFileOutputStream.close(EditLogFileOutputStream.java:157) > at org.apache.hadoop.hdfs.server.namenode.JournalSet$ > JournalAndStream.closeStream(JournalSet.java:115) > at org.apache.hadoop.hdfs.server.namenode.JournalSet$4.apply( > JournalSet.java:235) > at org.apache.hadoop.hdfs.server.namenode.JournalSet. > mapJournalsAndReportErrors(JournalSet.java:393) > at org.apache.hadoop.hdfs.server.namenode.JournalSet. > finalizeLogSegment(JournalSet.java:231) > at org.apache.hadoop.hdfs.server.namenode.FSEditLog. > endCurrentLogSegment(FSEditLog.java:1274) > - locked <0x00007f60487f17f8> (a org.apache.hadoop.hdfs.server. > namenode.FSEditLog) > at org.apache.hadoop.hdfs.server.namenode.FSEditLog. > rollEditLog(FSEditLog.java:1203) > - locked <0x00007f60487f17f8> (a org.apache.hadoop.hdfs.server. > namenode.FSEditLog) > at org.apache.hadoop.hdfs.server.namenode.FSImage.rollEditLog( > FSImage.java:1294) > at org.apache.hadoop.hdfs.server.namenode.FSNamesystem. > rollEditLog(FSNamesystem.java:5832) > at org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer. > rollEditLog(NameNodeRpcServer.java:1121) > at org.apache.hadoop.hdfs.protocolPB. > NamenodeProtocolServerSideTran > slatorPB.rollEditLog(NamenodeProtocolServerSideTranslatorPB.java:142) > at org.apache.hadoop.hdfs.protocol.proto.NamenodeProtocolProtos$ > NamenodeProtocolService$2.callBlockingMethod(NamenodeProtocolProtos.java: > 12025) > at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ > ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:616) > at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:969) > at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2049) > at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2045) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:422) > at org.apache.hadoop.security.UserGroupInformation.doAs( > UserGroupInformation.java:1657) > at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2045) > > >
