[
https://issues.apache.org/jira/browse/BIGTOP-952?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13912170#comment-13912170
]
jay vyas commented on BIGTOP-952:
---------------------------------
Well, after some more thought, I guess, it doesnt need to be a jar. I might be
overthinking this :)
I guess we can just do this, and then call this groovy file in the same place
that we used to call init-hdfs.sh .
{noformat}
#!/bin/bash
export HADOOP_CONF_DIR=/etc/hadoop/conf/
//usr/bin/env groovy -cp /usr/lib/hadoop/lib/*jar "$0" $@; exit $?
//.... romans code goes here,
{noformat}
right?
> init-hdfs.sh is dog slow. Let's replace it with a single VM call and better
> layout management
> ---------------------------------------------------------------------------------------------
>
> Key: BIGTOP-952
> URL: https://issues.apache.org/jira/browse/BIGTOP-952
> Project: Bigtop
> Issue Type: Improvement
> Components: Deployment
> Affects Versions: 0.5.0
> Reporter: Konstantin Boudnik
> Assignee: Roman Shaposhnik
> Priority: Blocker
> Fix For: 0.8.0
>
> Attachments: untar.groovy
>
>
> As has been proposed in [this
> patch|https://issues.apache.org/jira/secure/attachment/12575644/untarHdfs.groovy]
> by [~rvs] there's a very efficient way of creating layout in HDFS using a
> tarfile and Groovy script with direct call into DFS APIs.
> Let's making it happen.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
