[jira] [Commented] (DRILL-996) Build - allow other versions of Hadoop to be specified on command-line

["Jason Altekruse (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/DRILL-996?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14621486#comment-14621486 ] 

Jason Altekruse commented on DRILL-996:
---------------------------------------

We need to discuss this more before implementing it if it is really needed. While there are a few places in the pom files where we are declaring the version numbers as maven properties/variables, these have been used to removed redundancy, not to enable custom versions to the passed on the command line. We have not historically supported multiple versions of any our dependencies at the same time. If we need to start doing this, I think all of the supported versions need to be included in a multi-phase build where all unit tests will be run against each configuration.

> Build - allow other versions of Hadoop to be specified on command-line
> ----------------------------------------------------------------------
>
>                 Key: DRILL-996
>                 URL: https://issues.apache.org/jira/browse/DRILL-996
>             Project: Apache Drill
>          Issue Type: Improvement
>          Components: Tools, Build & Test
>            Reporter: Patrick Wong
>            Assignee: Jason Altekruse
>            Priority: Minor
>             Fix For: Future
>
>         Attachments: DRILL-996.1.patch.txt, DRILL-996.2.patch.txt
>
>
> Right now, in order to change the version of Hadoop that Drill is built against, you have to edit the POM. However, for automated build systems that build Drill against a variety of possible Hadoop versions, it would be much cleaner to be able to specify it on the command-line.



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