hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ahmed Radwan (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-3415) improve MiniMRYarnCluster & DistributedShell JAR resolution
Date Thu, 17 Nov 2011 11:32:51 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-3415?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13152000#comment-13152000
] 

Ahmed Radwan commented on MAPREDUCE-3415:
-----------------------------------------

Thanks Tucu! I have tested the patch and it works great. +1
I have a comment regrading clean up of created jar files. 
The files are now created in the temporary system directory, and the reliance will be on the
default system clean up (e.g. on reboot). I think this may cause problems on test dedicated
machines that will then require periodic cleanup of the tmp dir. After few minutes of testing
I have a dozen of jar files sitting in my tmp dir. I think it may be better to create these
jars in a maven predefined directory, that can be cleaned up through maven at the end. What
do you think?


                
> improve MiniMRYarnCluster & DistributedShell JAR resolution 
> ------------------------------------------------------------
>
>                 Key: MAPREDUCE-3415
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-3415
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: mrv2
>    Affects Versions: 0.24.0, 0.23.1
>            Reporter: Alejandro Abdelnur
>            Assignee: Alejandro Abdelnur
>             Fix For: 0.24.0, 0.23.1
>
>         Attachments: MAPREDUCE-3415.patch, MAPREDUCE-3415.patch, MAPREDUCE-3415.patch
>
>
> Current JAR resolution assumes the following:
> # The class used for JAR lookup is effectively in a JAR
> # A System property is set for testing with the location of the JAR
> The problem with #1 is that in some cases (when using the class in the same Maven module
where the class is, the class is not in a JAR but in a directory 'target/test-classes').
> The problem with #2 is the JAR does not exists at the time of running the test (packaging
comes after test and we are not doing integration testing yet thus won't work)
> In addition, this is required for streaming testcases, to have the JAR with streaming
classes for testing.

--
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