Mailing-List: contact mapreduce-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: mapreduce-issues@hadoop.apache.org
Date: Thu, 6 Sep 2012 16:22:12 +1100 (NCT)
From: "Luke Lu (JIRA)" <jira@apache.org>
To: mapreduce-issues@hadoop.apache.org
Message-ID: <1471431571.43038.1346908932249.JavaMail.jiratomcat@arcas>
Subject: [jira] [Commented] (MAPREDUCE-1700) User supplied dependencies may
 conflict with MapReduce system JARs
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/MAPREDUCE-1700?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13449430#comment-13449430 ] 

Luke Lu commented on MAPREDUCE-1700:
------------------------------------

bq. Without a user doing classloader gymnasitics and fancy packaging themselves, there is not always a way.

That's an interesting way to say that except for some ways that would always work, there is not always a way. Using the standard task API to bootstrap an OSGi container is reasonably straight forward :)

bq. A user cannot simply package a jar up and ask hadoop to execute it and expose to the user's execution environment only the public Hadoop API.

I do agree that there is a usability issue for certain (and arguably less common) use cases, where a user wants to use dependencies that conflict with client framework. However the proposed OSGi approach makes the usability worse for common cases: You'll always need OSGi bundles, which is a form of "fancy packaging", to run your jobs.

A more reasonable (and less heavy) solution would not require users to make any change (including adding metadata to their jars) to their existing code.
                
> User supplied dependencies may conflict with MapReduce system JARs
> ------------------------------------------------------------------
>
>                 Key: MAPREDUCE-1700
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1700
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: task
>            Reporter: Tom White
>            Assignee: Tom White
>         Attachments: MAPREDUCE-1700.patch, MAPREDUCE-1700.patch
>
>
> If user code has a dependency on a version of a JAR that is different to the one that happens to be used by Hadoop, then it may not work correctly. This happened with user code using a different version of Avro, as reported [here|https://issues.apache.org/jira/browse/AVRO-493?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12852081#action_12852081].
> The problem is analogous to the one that application servers have with WAR loading. Using a specialized classloader in the Child JVM is probably the way to solve this.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira