Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Flume, mail # dev - Patches for ElasticSearchSink


+
Edward Sargisson 2013-04-05, 19:14
+
Israel Ekpo 2013-04-05, 23:52
+
Mike Percy 2013-04-10, 20:41
Copy link to this message
-
Re: Patches for ElasticSearchSink
Israel Ekpo 2013-04-10, 21:26
I will update this wiki later next week with these steps as well as your
recommendations/additions.

It is one of the things on my todo list.

I also plan to add information on how new comers can locally or remotely
debug Apache Flume while it is still running using Eclipse IDE.

It is my hope that this information will help developers that might have
the time to help fix and resolve issues but do not necessary know what is
necessary to set up the environment needed to perform the task.

It think with time, as our documentation grows, finding information on how
things work within the project will become simpler and easier.

On 10 April 2013 16:41, Mike Percy <[EMAIL PROTECTED]> wrote:

> Thanks Israel for the quick follow up, minor suggestion to the very helpful
> instructions above: just assign group Flume as the reviewer, in most cases
> individuals would be on that group anyway (it just sends to the dev list)
>
> Certainly interested in improvements to the elastic search sink. Thanks
> Edward!
>
> Regards,
> Mike
>
>
>
> On Fri, Apr 5, 2013 at 4:52 PM, Israel Ekpo <[EMAIL PROTECTED]> wrote:
>
> > Hi Edward,
> >
> > It's good that you have patches available that resolves the issue.
> >
> > Here is what I would recommend to get the ball rolling, if you have some
> > time to spare:
> >
> > (1) Get the latest code from GIT
> >
> > git clone http://git-wip-us.apache.org/repos/asf/flume.git flume
> >
> > cd flume
> >
> > git checkout flume-1.4
> >
> > (2) Apply your changes to the code you have checked out
> >
> > (3) Try to compile the code changes, create a tarball and run the unit
> > tests to make sure that all is well with your changes
> >
> > mvn clean install
> >
> > (4) If all is well, then create the patches:
> >
> > git diff --no-prefix > /path/to/patch-file.patch
> >
> > It is very important to test the changes to make sure the work
> > satisfactorily before submitting the patch
> >
> > So you can configure the generated tarball and run it to see if it works
> > for you first
> >
> > Use the JIRA issue ID for the patch files you generate
> >
> > (5) Upload the patche file(s) to the JIRA issue and set the status to
> PATCH
> > AVAILABLE.
> >
> > (6) Create a review for your patch files here:
> >
> > https://reviews.apache.org/
> >
> > Pick "flume-git" as the repository and "Flume" as the group
> >
> > You can assign it to juhanic, mpercy, hshreedharan, brock or will
> >
> > (7) You can take a look at other code reviews here for some examples:
> >
> > https://reviews.apache.org/groups/Flume/
> >
> > On Fri, Apr 5, 2013 at 3:14 PM, Edward Sargisson <[EMAIL PROTECTED]>
> wrote:
> >
> > > Hi,
> > > I'm wondering how to work with somebody to get some patches for the
> > > ElasticSearchSink accepted into trunk.
> > >
> > > One of my colleagues at my unnamed employer wrote and submitted the
> > > original version. Since then, I have discovered and fixed a major
> defect
> > > around the name of the index to write to (FLUME-1782) and have just
> > > submitted an enhancement that allows users to calculate IDs for each
> > > document (FLUME-1972).
> > >
> > > I originally submitted FLUME-1782 on 13 December. There have been some
> > > questions raised but the patch(es) have yet to be accepted.
> > >
> > > I would really prefer these fixes to make it into 1.4.0 whenever it's
> > > released. Is there a way we can work together to resolve any problems?
> > >
> > > Cheers,
> > > Edward
> > >
> >
>