[
https://issues.apache.org/jira/browse/SQOOP-1266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13869623#comment-13869623
]
Sean Mackrory commented on SQOOP-1266:
--------------------------------------
+1 (non-committer). From an Apache Bigtop perspective, I like the idea of the
top-level commands specifying the major Sqoop 2 version. There have been
requests to distribute Sqoop 1.x and Sqoop 2.x concurrently, and this will make
it easier, clearer, and more consistent if a user ends up having both installed
side-by-side. Might I also suggest using "exec" for the commands you invoke as
the last line of these wrappers, though? Otherwise you end up with an extra
process in the tree without a very good reason - if a user is trying to track
down the process doing the actual work it's one less step to work out.
> Sqoop2: Introduce top level commands
> ------------------------------------
>
> Key: SQOOP-1266
> URL: https://issues.apache.org/jira/browse/SQOOP-1266
> Project: Sqoop
> Issue Type: Task
> Affects Versions: 1.99.3
> Reporter: Jarek Jarcec Cecho
> Assignee: Jarek Jarcec Cecho
> Fix For: 1.99.4
>
> Attachments: SQOOP-1266.patch
>
>
> Currently we have in Sqoop 2 one command line utility {{sqoop.sh}} that is
> user suppose to run with various arguments. As this shell script is not meant
> to be on the {{$PATH}}, we have full path to the script on all documentation
> pages. This very confusing for downstream consumption, as distributions such
> as BigTop are not exposing {{sqoop.sh}} script directly and user is confused
> how to apply the instructions in the docs.
> I would like to propose top level commands similarly as we currently have in
> Sqoop 1 and use them in the documentation. This will lead to much cleaner
> documentation and as a side effect our docs will be applicable to downstream
> distributions as well.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
