You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@zeppelin.apache.org by Woonsan Ko <wo...@apache.org> on 2016/05/11 20:42:30 UTC

Is it possible or feasible to deploy zeppelin as war?

Hi,

Maybe I'm asking about kind of corner case here, but is it possible or
feasible in the future to support packaging/deploying zeppelin as a
war?
In my specific use case (I'm working in web content management area),
I might want to be able to use zeppelin as a simple analytic platform
against JCR content data (possibly using jdbc interpreter for jcr). I
assume there's no big overhead by packaging/using it as war on servlet
container if it's connecting to a remote sparks.
Would it be doable or is it a too narrow use case to support war
packaging in zeppelin?

Thanks in advance,

Woonsan

Re: Is it possible or feasible to deploy zeppelin as war?

Posted by Woonsan Ko <wo...@apache.org>.
Hi moon,

Thanks a lot for the remark. I'll keep that in mind.

Cheers,

Woonsan


On Mon, May 16, 2016 at 9:38 PM, moon soo Lee <mo...@apache.org> wrote:
> Hi Woosan,
>
> Currently, Zeppelin starts interpreter as a separate process using
> bin/interpreter.sh, and that process need to add one of subdirectory under
> './interpreter' directory into classpath.
>
> I guess packaging everything into war will not be simple unless change
> current interpreter starting mechanism.
>
> Hope this helps.
>
> Thanks,
> moon
>
> On Wed, May 11, 2016 at 1:42 PM Woonsan Ko <wo...@apache.org> wrote:
>>
>> Hi,
>>
>> Maybe I'm asking about kind of corner case here, but is it possible or
>> feasible in the future to support packaging/deploying zeppelin as a
>> war?
>> In my specific use case (I'm working in web content management area),
>> I might want to be able to use zeppelin as a simple analytic platform
>> against JCR content data (possibly using jdbc interpreter for jcr). I
>> assume there's no big overhead by packaging/using it as war on servlet
>> container if it's connecting to a remote sparks.
>> Would it be doable or is it a too narrow use case to support war
>> packaging in zeppelin?
>>
>> Thanks in advance,
>>
>> Woonsan

Re: Is it possible or feasible to deploy zeppelin as war?

Posted by moon soo Lee <mo...@apache.org>.
Hi Woosan,

Currently, Zeppelin starts interpreter as a separate process using
bin/interpreter.sh, and that process need to add one of subdirectory under
'./interpreter' directory into classpath.

I guess packaging everything into war will not be simple unless change
current interpreter starting mechanism.

Hope this helps.

Thanks,
moon

On Wed, May 11, 2016 at 1:42 PM Woonsan Ko <wo...@apache.org> wrote:

> Hi,
>
> Maybe I'm asking about kind of corner case here, but is it possible or
> feasible in the future to support packaging/deploying zeppelin as a
> war?
> In my specific use case (I'm working in web content management area),
> I might want to be able to use zeppelin as a simple analytic platform
> against JCR content data (possibly using jdbc interpreter for jcr). I
> assume there's no big overhead by packaging/using it as war on servlet
> container if it's connecting to a remote sparks.
> Would it be doable or is it a too narrow use case to support war
> packaging in zeppelin?
>
> Thanks in advance,
>
> Woonsan
>