sandflee created YARN-4426:
------------------------------
Summary: unhealthy disk makes NM LOST
Key: YARN-4426
URL: https://issues.apache.org/jira/browse/YARN-4426
Project: Hadoop YARN
Issue Type: Bug
Reporter: sandflee
nm are hanged because mkdir hangs in DiskHealthMonitor-Timer, and
nodeStatusUpdater couldn't get sync lock in getNodeStatus
"DiskHealthMonitor-Timer" daemon prio=10 tid=0x00007f4b3d867000 nid=0x50c8
runnable [0x00007f4b27ef9000]
java.lang.Thread.State: RUNNABLE
at java.io.UnixFileSystem.createDirectory(Native Method)
at java.io.File.mkdir(File.java:1310)
at
org.apache.hadoop.util.DiskChecker.mkdirsWithExistsCheck(DiskChecker.java:67)
at org.apache.hadoop.util.DiskChecker.checkDir(DiskChecker.java:90)
at
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection.verifyDirUsingMkdir(DirectoryCollection.java:338)
at
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection.testDirs(DirectoryCollection.java:310)
at
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection.checkDirs(DirectoryCollection.java:230)
- locked <0x00000000f8970408> (a
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection)
at
org.apache.hadoop.yarn.server.nodemanager.LocalDirsHandlerService.checkDirs(LocalDirsHandlerService.java:361)
at
org.apache.hadoop.yarn.server.nodemanager.LocalDirsHandlerService.access$400(LocalDirsHandlerService.java:51)
at
org.apache.hadoop.yarn.server.nodemanager.LocalDirsHandlerService$MonitoringTimerTask.run(LocalDirsHandlerService.java:123)
at java.util.TimerThread.mainLoop(Timer.java:555)
at java.util.TimerThread.run(Timer.java:505)
"Node Status Updater" prio=10 tid=0x00007f4b3cd6d800 nid=0x4af5 waiting for
monitor entry [0x00007f4b1c141000]
java.lang.Thread.State: BLOCKED (on object monitor)
at
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection.getFailedDirs(DirectoryCollection.java:170)
- waiting to lock <0x00000000f8970408> (a
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection)
at
org.apache.hadoop.yarn.server.nodemanager.LocalDirsHandlerService.getDisksHealthReport(LocalDirsHandlerService.java:259)
at
org.apache.hadoop.yarn.server.nodemanager.NodeHealthCheckerService.getHealthReport(NodeHealthCheckerService.java:58)
at
org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl.getNodeStatus(NodeStatusUpdaterImpl.java:365)
at
org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl.access$100(NodeStatusUpdaterImpl.java:77)
at
org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl$1.run(NodeStatusUpdaterImpl.java:588)
at java.lang.Thread.run(Thread.java:745)
"AsyncDispatcher event handler" prio=10 tid=0x00007f4b3da24000 nid=0x50d9
waiting for monitor entry [0x00007f4b245b6000]
java.lang.Thread.State: BLOCKED (on object monitor)
at
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection.getGoodDirs(DirectoryCollection.java:163)
- waiting to lock <0x00000000f8970408> (a
org.apache.hadoop.yarn.server.nodemanager.DirectoryCollection)
at
org.apache.hadoop.yarn.server.nodemanager.LocalDirsHandlerService.getLocalDirsForCleanup(LocalDirsHandlerService.java:229)
at
org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ResourceLocalizationService.handleCleanupContainerResources(ResourceLocalizationService.java:497)
at
org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ResourceLocalizationService.handle(ResourceLocalizationService.java:395)
at
org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ResourceLocalizationService.handle(ResourceLocalizationService.java:134)
at
org.apache.hadoop.yarn.event.AsyncDispatcher.dispatch(AsyncDispatcher.java:191)
at
org.apache.hadoop.yarn.event.AsyncDispatcher$1.run(AsyncDispatcher.java:124)
at java.lang.Thread.run(Thread.java:745)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
