hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "John Novatnack (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-2389) Spurious EOFExceptions reading SpillRecord index files
Date Wed, 04 May 2011 20:32:03 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-2389?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13028941#comment-13028941
] 

John Novatnack commented on MAPREDUCE-2389:
-------------------------------------------

Todd: Thanks we will try that.  Most of the time it seems to just be causing occasional task
failure, but we have also seen the exception happen on a handful of slaves simultaneously
and eventually push the jobtracker to run out of heap.  Right now our heap is 1gb for the
jobtracker with 90 slaves so we'll try increasing it and see if the problem persists with
that and the older version of Jetty.

OS: Ubuntu 10.10
JVM:
java version "1.6.0_24"
Java(TM) SE Runtime Environment (build 1.6.0_24-b07)

> Spurious EOFExceptions reading SpillRecord index files
> ------------------------------------------------------
>
>                 Key: MAPREDUCE-2389
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2389
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: tasktracker
>    Affects Versions: 0.22.0
>         Environment: Seen on RHEL 5.5, RHEL 6.0, local dirs on ext3, Java 6u20 and 6u24
>            Reporter: Todd Lipcon
>            Priority: Critical
>         Attachments: stap-output.txt
>
>
> In large jobs, I see around 1 shuffle fetch out of every million fetches fail with an
EOFException reading the SpillRecord index file. After lots of investigation, including systemtap,
it looks like the read() syscall is actually returning a premature "0" result for no reason,
so this is likely a kernel or filesystem bug which is exacerbated by some workload the TT
does.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message