hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anupam Seth (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-7993) Hadoop ignores old-style config options for enabling compressed output
Date Thu, 26 Jan 2012 17:58:38 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-7993?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13194026#comment-13194026
] 

Anupam Seth commented on HADOOP-7993:
-------------------------------------

Just to clarify the reason for proposing the code change the way it is in the patch is that
upon investigation, our understanding was that the old-style config was being written to file
by the JobSubmitter before the replacement with the keys listed in MR deprecated keys are
invoked, e.g. by a get. Thus, when the AppMaster tried to read in the config coupled together
with mapred-default which currently has this set to false, the old-style config was being
overriden to be whatever was specified in the default file. As a solution, therefore, we force
handling deprecation on all keys before the job.xml gets written out.
                
> Hadoop ignores old-style config options for enabling compressed output
> ----------------------------------------------------------------------
>
>                 Key: HADOOP-7993
>                 URL: https://issues.apache.org/jira/browse/HADOOP-7993
>             Project: Hadoop Common
>          Issue Type: Bug
>          Components: conf
>    Affects Versions: 0.23.1
>            Reporter: Anupam Seth
>            Assignee: Anupam Seth
>         Attachments: HADOOP-7993_branch_0_23.patch, HADOOP-7993_branch_0_23.patch
>
>
> Hadoop seems to ignore the config options even though they are printed as deprecation
warnings in the log: mapred.output.compress and
> mapred.output.compression.codec
> - settings that work on 0.20 but not on 0.23
> mapred.output.compress=true
> mapred.output.compression.codec=org.apache.hadoop.io.compress.BZip2Codec
> - settings that work on 0.23
> mapreduce.output.fileoutputformat.compress=true
> mapreduce.output.fileoutputformat.compress.codec=org.apache.hadoop.io.compress.BZip2Codec
> This breaks backwards compatibility and causes existing jobs to fail.
> This was found to happen due to the JobSubmitter writing out the job.xml file with the
old-style configs and can be fixed by handdling deprecation before the file is written out.

--
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

        

Mime
View raw message