[YARN-1091] All containers localization fails in NM when any one of the configured nm local-dir disk becomes full

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Duplicate
• Affects Version/s: 2.0.5-alpha
• Component/s: nodemanager
• Assignee: Devaraj Kavali
• Reporter: Devaraj Kavali
• Created: 22/Aug/13 10:30
• Updated: 22/Aug/13 15:42
• Resolved: 22/Aug/13 15:42

Description

2013-08-22 13:54:22,100 WARN org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor: Unable to create app directory /opt/nish/usercache/nish/appcache/application_1377151891396_0017
java.io.IOException: mkdir of /opt/nish/usercache/nish/appcache/application_1377151891396_0017 failed
	at org.apache.hadoop.fs.FileSystem.primitiveMkdir(FileSystem.java:1125)
	at org.apache.hadoop.fs.DelegateToFileSystem.mkdir(DelegateToFileSystem.java:150)
	at org.apache.hadoop.fs.FilterFs.mkdir(FilterFs.java:187)
	at org.apache.hadoop.fs.FileContext$4.next(FileContext.java:730)
	at org.apache.hadoop.fs.FileContext$4.next(FileContext.java:726)
	at org.apache.hadoop.fs.FileContext$FSLinkResolver.resolve(FileContext.java:2379)
	at org.apache.hadoop.fs.FileContext.mkdir(FileContext.java:726)
	at org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor.createDir(DefaultContainerExecutor.java:330)
	at org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor.createAppDirs(DefaultContainerExecutor.java:426)
	at org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor.startLocalizer(DefaultContainerExecutor.java:90)
	at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ResourceLocalizationService$LocalizerRunner.run(ResourceLocalizationService.java:859)
2013-08-22 13:54:22,102 INFO org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor: Copying from /home/nish/new/JAN_4/nmlocal/nmPrivate/container_1377151891396_0017_01_000263.tokens to /home/nish/new/JAN_4/nmlocal/usercache/nish/appcache/application_1377151891396_0017/container_1377151891396_0017_01_000263.tokens
2013-08-22 13:54:22,102 INFO org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor: CWD set to /home/nish/new/JAN_4/nmlocal/usercache/nish/appcache/application_1377151891396_0017 = file:/home/nish/new/JAN_4/nmlocal/usercache/nish/appcache/application_1377151891396_0017
2013-08-22 13:54:22,103 INFO org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ResourceLocalizationService: Localizer failed
java.io.FileNotFoundException: File file:/opt/nish/usercache/nish/appcache/application_1377151891396_0017 does not exist
	at org.apache.hadoop.fs.RawLocalFileSystem.getFileStatus(RawLocalFileSystem.java:492)
	at org.apache.hadoop.fs.FileSystem.primitiveMkdir(FileSystem.java:1112)
	at org.apache.hadoop.fs.DelegateToFileSystem.mkdir(DelegateToFileSystem.java:150)
	at org.apache.hadoop.fs.FilterFs.mkdir(FilterFs.java:187)
	at org.apache.hadoop.fs.FileContext$4.next(FileContext.java:730)
	at org.apache.hadoop.fs.FileContext$4.next(FileContext.java:726)
	at org.apache.hadoop.fs.FileContext$FSLinkResolver.resolve(FileContext.java:2379)
	at org.apache.hadoop.fs.FileContext.mkdir(FileContext.java:726)
	at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs(ContainerLocalizer.java:391)
	at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.runLocalization(ContainerLocalizer.java:130)
	at org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor.startLocalizer(DefaultContainerExecutor.java:103)
	at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ResourceLocalizationService$LocalizerRunner.run(ResourceLocalizationService.java:859)
2013-08-22 13:54:22,104 INFO org.apache.hadoop.yarn.server.nodemanager.containermanager.container.Container: Container container_1377151891396_0017_01_000263 transitioned from LOCALIZING to LOCALIZATION_FAILED

Issue Links

• duplicates: YARN-257 NM should gracefully handle a full local disk