Interpreter temporary files

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Interpreter temporary files

MUEDSAM, JOHAN
We’ve run into an issue where the Spark and R interpreters creates temporary files with static names, /tmp/zeppelin_pyspark.py and /tmp/zeppelin_sparkr.R respectively. The reason it doesn’t work for us is that we’re running multiple instances on the same server (multi-user behind nginx reverse proxy) and we get ownership collisions of the files between the instances.

I’ve patched the two interpreters locally using java.io.createTempFile instead of the static file names. Would it be of use for me to create a pull request with this fix? Or should I just open a JIRA?

Thanks,
Johan



Reply | Threaded
Open this post in threaded view
|

Re: Interpreter temporary files

moon
Administrator
Hi Johan,

Thanks for considering the contribution.
Creating jira issue is not mandatory in this project. But it's always good
idea to open an JIRA issue for a pullrequest.

JIRA issue is not only a useful for the discussion but also useful to track
extra informations such as fix version. Moreover, release note includes
list of issues has been resolved, generated by JIRA.

Thanks,
moon

On Thu, Apr 7, 2016 at 7:04 AM MUEDSAM, JOHAN <[hidden email]> wrote:

> We’ve run into an issue where the Spark and R interpreters creates
> temporary files with static names, /tmp/zeppelin_pyspark.py and
> /tmp/zeppelin_sparkr.R respectively. The reason it doesn’t work for us is
> that we’re running multiple instances on the same server (multi-user behind
> nginx reverse proxy) and we get ownership collisions of the files between
> the instances.
>
> I’ve patched the two interpreters locally using java.io.createTempFile
> instead of the static file names. Would it be of use for me to create a
> pull request with this fix? Or should I just open a JIRA?
>
> Thanks,
> Johan
>
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: Interpreter temporary files

MUEDSAM, JOHAN
Thanks moon, I created pull request #819 for this - https://github.com/apache/incubator-zeppelin/pull/819
Unfortunately no Jira, realized I don't have an account and didn’t have time to set one up for now.

/Johan






On 4/6/16, 6:20 PM, "moon soo Lee" <[hidden email]> wrote:

>Hi Johan,
>
>Thanks for considering the contribution.
>Creating jira issue is not mandatory in this project. But it's always good
>idea to open an JIRA issue for a pullrequest.
>
>JIRA issue is not only a useful for the discussion but also useful to track
>extra informations such as fix version. Moreover, release note includes
>list of issues has been resolved, generated by JIRA.
>
>Thanks,
>moon
>
>On Thu, Apr 7, 2016 at 7:04 AM MUEDSAM, JOHAN <[hidden email]> wrote:
>
>> We’ve run into an issue where the Spark and R interpreters creates
>> temporary files with static names, /tmp/zeppelin_pyspark.py and
>> /tmp/zeppelin_sparkr.R respectively. The reason it doesn’t work for us is
>> that we’re running multiple instances on the same server (multi-user behind
>> nginx reverse proxy) and we get ownership collisions of the files between
>> the instances.
>>
>> I’ve patched the two interpreters locally using java.io.createTempFile
>> instead of the static file names. Would it be of use for me to create a
>> pull request with this fix? Or should I just open a JIRA?
>>
>> Thanks,
>> Johan
>>
>>
>>
>>
Reply | Threaded
Open this post in threaded view
|

Re: Interpreter temporary files

MUEDSAM, JOHAN
Need more caffeine, it’s PR 823, not 819, my bad..

/Johan




On 4/7/16, 12:08 AM, "MUEDSAM, JOHAN" <[hidden email]> wrote:


>Thanks moon, I created pull request #819 for this - https://github.com/apache/incubator-zeppelin/pull/819
>Unfortunately no Jira, realized I don't have an account and didn’t have time to set one up for now.
>
>/Johan
>
>
>
>
>
>
>On 4/6/16, 6:20 PM, "moon soo Lee" <[hidden email]> wrote:
>
>>Hi Johan,
>>
>>Thanks for considering the contribution.
>>Creating jira issue is not mandatory in this project. But it's always good
>>idea to open an JIRA issue for a pullrequest.
>>
>>JIRA issue is not only a useful for the discussion but also useful to track
>>extra informations such as fix version. Moreover, release note includes
>>list of issues has been resolved, generated by JIRA.
>>
>>Thanks,
>>moon
>>
>>On Thu, Apr 7, 2016 at 7:04 AM MUEDSAM, JOHAN <[hidden email]> wrote:
>>
>>> We’ve run into an issue where the Spark and R interpreters creates
>>> temporary files with static names, /tmp/zeppelin_pyspark.py and
>>> /tmp/zeppelin_sparkr.R respectively. The reason it doesn’t work for us is
>>> that we’re running multiple instances on the same server (multi-user behind
>>> nginx reverse proxy) and we get ownership collisions of the files between
>>> the instances.
>>>
>>> I’ve patched the two interpreters locally using java.io.createTempFile
>>> instead of the static file names. Would it be of use for me to create a
>>> pull request with this fix? Or should I just open a JIRA?
>>>
>>> Thanks,
>>> Johan
>>>
>>>
>>>
>>>