How to add additional dependencies to Hadoop slaves

[Konstantinos Tsakalozos]

Hi Merlijn,

I do not think there is a way to deploy pip and apt dependencies at the
moment on the slaves. We have seen a similar need on the charms acting as
clients to a hadoop cluster (eg hadoop-client). Users often need to install
libraries (usually) in the form of jars and include them in job submission.
We were thinking of a "generic dependency install" relation for this
purpose. This relation might make sense to also be present on the slaves.

Would you be able to give us some more insight on what dependencies you are
interested in deploying on the slaves? Needless to say we are welcoming any
PR you may have.

Thanks,
Konstantinos

PS. Sorry for late response most of us are on vacations this week.






On Tue, Dec 20, 2016 at 6:24 PM, Merlijn Sebrechts <
merlijn.sebrechts at gmail.com> wrote:

> Hi all
>
>
> What's the best way to add additional dependencies (pip and apt) to Hadoop
> slaves? Is there a way to request additional dependencies over
> relationships? Or do I just make an "additional dependencies" subordinate
> charm?
>
>
>
> Kind regards
> Merlijn Sebrechts
>
>
>
>
>
>
> --
> Bigdata mailing list
> Bigdata at lists.ubuntu.com
> Modify settings or unsubscribe at: https://lists.ubuntu.com/
> mailman/listinfo/bigdata
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/bigdata/attachments/20161222/b9ea03c0/attachment.html>