[YARN-527] Local filecache mkdir fails

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Duplicate
• Affects Version/s: 2.0.0-alpha
• Component/s: nodemanager

Environment

RHEL 6.3 with CDH4.1.3 Hadoop, HA with two name nodes and six worker nodes.

• Assignee: Unassigned
• Reporter: Knut O. Hellan
• Created: 02/Apr/13 08:54
• Updated: 04/Apr/13 17:12
• Resolved: 04/Apr/13 17:12

Description

Jobs failed with no other explanation than this stack trace:

2013-03-29 16:46:02,671 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Diag
nostics report from attempt_1364591875320_0017_m_000000_0: java.io.IOException: mkdir of /disk3/yarn/local/filecache/-42307893
55400878397 failed
at org.apache.hadoop.fs.FileSystem.primitiveMkdir(FileSystem.java:932)
at org.apache.hadoop.fs.DelegateToFileSystem.mkdir(DelegateToFileSystem.java:143)
at org.apache.hadoop.fs.FilterFs.mkdir(FilterFs.java:189)
at org.apache.hadoop.fs.FileContext$4.next(FileContext.java:706)
at org.apache.hadoop.fs.FileContext$4.next(FileContext.java:703)
at org.apache.hadoop.fs.FileContext$FSLinkResolver.resolve(FileContext.java:2333)
at org.apache.hadoop.fs.FileContext.mkdir(FileContext.java:703)
at org.apache.hadoop.yarn.util.FSDownload.call(FSDownload.java:147)
at org.apache.hadoop.yarn.util.FSDownload.call(FSDownload.java:49)
at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
at java.util.concurrent.FutureTask.run(FutureTask.java:138)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441)
at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
at java.util.concurrent.FutureTask.run(FutureTask.java:138)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:662)

Manually creating the directory worked. This behavior was common to at least several nodes in the cluster.

The situation was resolved by removing and recreating all /disk?/yarn/local/filecache directories on all nodes.

It is unclear whether Yarn struggled with the number of files or if there were corrupt files in the caches. The situation was triggered by a node dying.

Issue Links

• is duplicated by: YARN-467 Jobs fail during resource localization when public distributed-cache hits unix directory limits