HUE... spark, hive, zeppelin

[Andrew Mcleod]

Right now, nothing - I had a dummy charm and dummy interface when I was
building locally, but there is no layered oozie charm. For testing i
deployed the existing oozie charm from the charm store and configured it
manually.



On Fri, Feb 26, 2016 at 6:01 PM, Merlijn Sebrechts <
merlijn.sebrechts at gmail.com> wrote:

> Hi Andrew
>
>
> What Charm are you using for oozie at the moment?
>
>
>
> Kind regards
> Merlijn
>
> 2016-02-26 16:45 GMT+01:00 Andrew Mcleod <andrew.mcleod at canonical.com>:
>
>> Hey Merlijn,
>>
>> That sounds about right,  but you don't have to change layers.yaml to use
>> local layers - if the layer exists in your JUJU_LAYERS directory, it will
>> use that by default (as long as its named correctly, so for example,
>> layer-hadoop-base would need to be in a directory called hadoop-base -
>> exclude the 'layer' prefix if it has one).
>>
>> Andrew
>>
>> On Fri, Feb 26, 2016 at 4:38 PM, Merlijn Sebrechts <
>> merlijn.sebrechts at gmail.com> wrote:
>>
>>> Hi all
>>>
>>>
>>> I'm currently reviewing these changes. I'd like to deploy these changes.
>>> My current approach:
>>>
>>> - clone repo's
>>> - merge PR's
>>> - change layers.yaml to use local layers
>>> - Deploy charm
>>>
>>> Is there a less cumbersome way to do this?
>>>
>>>
>>>
>>> Kind regards
>>> Merlijn
>>>
>>> 2016-02-18 15:31 GMT+01:00 Cory Johns <cory.johns at canonical.com>:
>>>
>>>> (Adding the bigdata list.)
>>>>
>>>> Thanks, Andrew.  I'd also like to add my own small PR:
>>>>
>>>> https://github.com/juju-solutions/interface-spark/pull/3
>>>>
>>>> Merlijn, I know you were interested in and perhaps working on Hue
>>>> yourself.  Hopefully there hasn't been too much parallel effort going on,
>>>> and hopefully this work benefits you.  Perhaps you could give it a review
>>>> as well?
>>>>
>>>> On Thu, Feb 18, 2016 at 7:25 AM, Andrew Mcleod <
>>>> andrew.mcleod at canonical.com> wrote:
>>>>
>>>>> Ok, so I'm done for HUE for the time being.
>>>>>
>>>>> !!!!!!! PLEASE REVIEW !!!!!!!
>>>>>
>>>>> https://github.com/juju-solutions/layer-apache-zeppelin/pull/3
>>>>> https://github.com/juju-solutions/layer-apache-spark/pull/7
>>>>> https://github.com/juju-solutions/layer-apache-hive/pull/8
>>>>> https://github.com/juju-solutions/interface-spark/pull/2
>>>>>
>>>>>
>>>>> I have changed some states (from related => joined and available =>
>>>>> ready) in the spark interface and spark reactive (as discussed)
>>>>>
>>>>>
>>>>>
>>>>> *** benchmark layer / interface will need to be updated to join to
>>>>> this new spark layer*
>>>>>
>>>>> I've added the livy server into spark with some caveats:
>>>>>
>>>>> 1. The start and stop clauses are a bit dodgy (i dont know of a better
>>>>> way to do it right now, and the jps process name is "Main"...)
>>>>>
>>>>> 2. livy CLASSPATH is hard coded - `hadoop classpath` should return the
>>>>> right value, however i couldnt use the utils.env as it was putting a lone
>>>>> ‘=‘ after the classpath insertion, i.e /etc/environments looks like this:
>>>>>
>>>>> STUFF="thing"
>>>>> OTHERSTUFF="otherthing"
>>>>> CLASSPATH="/etc/dir:/dir:/dir:/dir............/dir/dir/dir"
>>>>> =
>>>>>
>>>>> ^ Equals sign out of nowhere.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>> --
>>>> 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/20160226/a96a0742/attachment-0001.html>