[
https://issues.apache.org/jira/browse/TEZ-1300?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14072584#comment-14072584
]
Siddharth Seth commented on TEZ-1300:
-------------------------------------
bq. If possible, let us use only one. Having 2 different types will be very
confusing for users. Similar comment for partial tarballs - can we figure out a
way to steer users more towards the recommended option - maybe document the
second option in an addendum after the initial guide?
That can be done. Your concern about multiple copies of Hadoop on the classpath
however still exists. Maybe we need to add an explicit tez command line.
bq. Not sure why option 3 is needed. Shouldn't doing a -Dtar on option 1 be
enough? Are the structures of the 2 tarballs different?
Target 3 is to always generate a .tar.gz - which is the recommended install.
Option 1 and Option 3 are otherwise identical. Do we need the option of
generating a dir / tarball - or should we just generate .tar.gz files by
default.
> Change default tez classpath to not include hadoop jars from the cluster
> ------------------------------------------------------------------------
>
> Key: TEZ-1300
> URL: https://issues.apache.org/jira/browse/TEZ-1300
> Project: Apache Tez
> Issue Type: Improvement
> Reporter: Siddharth Seth
> Assignee: Siddharth Seth
> Attachments: TEZ-1300.1.txt
>
>
--
This message was sent by Atlassian JIRA
(v6.2#6252)
