hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Robert Chansler (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-4362) Hadoop Streaming failed with large number of input files
Date Wed, 08 Oct 2008 18:03:44 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-4362?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12638008#action_12638008
] 

Robert Chansler commented on HADOOP-4362:
-----------------------------------------

Yes! The good news this morning is that Hairong now has a good understanding
of the issue.






> Hadoop Streaming failed with large number of input files
> --------------------------------------------------------
>
>                 Key: HADOOP-4362
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4362
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: dfs
>    Affects Versions: 0.18.1
>            Reporter: Peeyush Bishnoi
>            Assignee: Robert Chansler
>            Priority: Blocker
>             Fix For: 0.18.2, 0.19.0
>
>
> Simple job failed with "java.lang.ArrayIndexOutOfBoundsException" when the mapper is
/bin/cat and the number of input files is large.
> $  hadoop jar $HADOOP_HOME/hadoop-streaming.jar -input in_data -output op_data -mapper
/bin/cat -reducer NONE
> additionalConfSpec_:null
> null=@@@userJobConfProps_.get(stream.shipped.hadoopstreaming
> packageJobJar: [/tmp/hadoop-unjar49637/] []
> /tmp/streamjob49638.jar tmpDir=/tmp
> 08/10/07 07:03:09 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments.
Applications should
> implement Tool for the same.
> 08/10/07 07:03:11 INFO mapred.FileInputFormat: Total input paths to process : 16365
> 08/10/07 07:03:12 INFO mapred.FileInputFormat: Total input paths to process : 16365
> 08/10/07 07:03:15 ERROR streaming.StreamJob: Error Launching job : java.io.IOException:
> java.lang.ArrayIndexOutOfBoundsException
> Streaming Job Failed!
> But when the input number of files are less job does not fail . 
> $ hadoop  jar $HADOOP_HOME/hadoop-streaming.jar -input inp_data1 -output op_data1 -mapper
/bin/cat -reducer NONE
> additionalConfSpec_:null
> null=@@@userJobConfProps_.get(stream.shipped.hadoopstreaming
> packageJobJar: [/tmp/hadoop-unjar3725/] []
> /tmp/streamjob3726.jar tmpDir=/tmp
> 08/10/07 07:06:37 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments.
Applications should
> implement Tool for the same.
> 08/10/07 07:06:39 INFO mapred.FileInputFormat: Total input paths to process : 16
> 08/10/07 07:06:39 INFO mapred.FileInputFormat: Total input paths to process : 16
> 08/10/07 07:06:42 INFO streaming.StreamJob: getLocalDirs():
> [/var/mapred/local]
> 08/10/07 07:06:42 INFO streaming.StreamJob: Running job: job_200810070645_0006
> 08/10/07 07:06:42 INFO streaming.StreamJob: To kill this job, run:
> 08/10/07 07:06:42 INFO streaming.StreamJob: hadoop job -Dmapred.job.tracker=login1:51981
-kill job_200810070645_0006
> 08/10/07 07:06:42 INFO streaming.StreamJob: Tracking URL: http://login1:52941/jobdetails.jsp?jobid=job_200810070645_0006
> 08/10/07 07:06:43 INFO streaming.StreamJob:  map 0%  reduce 0%
> 08/10/07 07:06:46 INFO streaming.StreamJob:  map 44%  reduce 0%
> 08/10/07 07:06:47 INFO streaming.StreamJob:  map 75%  reduce 0%
> 08/10/07 07:06:48 INFO streaming.StreamJob:  map 88%  reduce 0%
> 08/10/07 07:06:49 INFO streaming.StreamJob:  map 100%  reduce 100%
> 08/10/07 07:06:49 INFO streaming.StreamJob: Job complete: job_200810070645_0006
> 08/10/07 07:06:49 INFO streaming.StreamJob: Output: op_data1

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message