hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gera Shegalov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-6793) io.sort.factor code default and mapred-default.xml values inconsistent
Date Thu, 27 Oct 2016 05:44:58 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-6793?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15610764#comment-15610764
] 

Gera Shegalov commented on MAPREDUCE-6793:
------------------------------------------

Hi [~Prabhu Joseph], thanks for working this JIRA. I suggest we don't change the effective
default in this JIRA since it requires a separate discussion, and has implication on the number
of random vs sequential iops + memory consumption. I would rather suggest just a cosmetic
change to introduce a constant DEFAULT_IO_SORT_FACTOR = 10 in MRJobConfig and change code
to 

.getInt(MRJobConfig.IO_SORT_FACTOR, MRJobConfig.DEFAULT_IO_SORT_FACTOR)

in MapTask.java and MergeManagerImpl.java

[~rohithsharma] is it ok with you?



> io.sort.factor code default and mapred-default.xml values inconsistent
> ----------------------------------------------------------------------
>
>                 Key: MAPREDUCE-6793
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-6793
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: task
>    Affects Versions: 2.6.5, 3.0.0-alpha1
>            Reporter: Gera Shegalov
>            Assignee: Prabhu Joseph
>            Priority: Trivial
>              Labels: noob
>         Attachments: 0001-MAPREDUCE-6793.patch
>
>
> The actual default value in mapred-default.xml:
> {code}
> <property>
>   <name>mapreduce.task.io.sort.factor</name>
>   <value>10</value>
>   <description>The number of streams to merge at once while sorting
>   files.  This determines the number of open file handles.</description>
> </property>
> {code}
> However, MapTask and MergeManagerImpl, are coded with:
> {code}       
>  int mergeFactor = job.getInt(JobContext.IO_SORT_FACTOR, 100);
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: mapreduce-issues-help@hadoop.apache.org


Mime
View raw message