hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew Wang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-3774) NPE in DFSInputStream.read() when called on non-existing block
Date Thu, 16 Aug 2012 02:14:38 GMT

    [ https://issues.apache.org/jira/browse/HDFS-3774?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13435719#comment-13435719

Andrew Wang commented on HDFS-3774:

I cleaned up what looked like duplicates of this issue (likely caused by the recent JIRA unhappiness).
Till, apologies if I mis-dup'd one of your JIRAs; I also put you back on the watch list for
this one.
> NPE in DFSInputStream.read() when called on non-existing block
> --------------------------------------------------------------
>                 Key: HDFS-3774
>                 URL: https://issues.apache.org/jira/browse/HDFS-3774
>             Project: Hadoop HDFS
>          Issue Type: Bug
>    Affects Versions: 2.0.0-alpha
>         Environment: CDH4.0.1 using yarn, log aggregation
>            Reporter: Till Machens
> Calling "yarn logs" on a *running* application results in a NPE:
> $ yarn logs -applicationId app_1344439655593_0002
> Exception in thread "main" java.lang.NullPointerException
> 	at org.apache.hadoop.hdfs.DFSInputStream.read(DFSInputStream.java:606)
> 	at java.io.DataInputStream.readFully(DataInputStream.java:178)
> 	at java.io.DataInputStream.readLong(DataInputStream.java:399)
> 	at org.apache.hadoop.io.file.tfile.BCFile$Reader.<init>(BCFile.java:623)
> 	at org.apache.hadoop.io.file.tfile.TFile$Reader.<init>(TFile.java:803)
> 	at org.apache.hadoop.yarn.logaggregation.AggregatedLogFormat$LogReader.<init>(AggregatedLogFormat.java:266)
> 	at org.apache.hadoop.yarn.logaggregation.LogDumper.dumpAllContainersLogs(LogDumper.java:209)
> 	at org.apache.hadoop.yarn.logaggregation.LogDumper.run(LogDumper.java:109)
> 	at org.apache.hadoop.yarn.logaggregation.LogDumper.main(LogDumper.java:242)
> Line numbers in the stack trace differ slightly from 2.0.0-alpha, due to the Cloudera
patches, but the bug seems to have its origin here (patch for HDFS-1371: https://svn.apache.org/repos/asf/hadoop/hdfs/trunk@1125145).
> At first sight this seems to be a two layer problem:
> - yarn logs -- designed for finished apps only :( -- is looking for non-existing files
here (?)
> - the used HDFS-implementation, which should report block replica corruptions here, but
results in NPE instead

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira


View raw message