Maybe you can use rest API to create a new interpreter and update the
interpreter setting
https://zeppelin.apache.org/docs/0.10.0/usage/rest_api/interpreter.html#create-a-new-interpreter-setting
https://zeppelin.apache.org/docs/0.10.0/usage/rest_api/interpreter.html#update-an-interpreter-setting
On Tue, Jan 11, 2022 at 9:25 PM Baptuste <baptiste.meyn...@gmail.com> wrote:
> Hello,
>
> I use Ansible playbooks to deploy Apache Zeppelin 0.10.0 on RedHat 8, running
> with java 1.8_301.
>
>
> To provide a Zeppelin solution for my team with all Interpreter
> configurations already set. Ansible change default value inside
> $ZEPPELIN_HOME/interpreter/{{INTERPRETER_NAME}}/interpreter-setting.jon (it's
> work well).
> On some case, for example for Cassandra interpreter we have multiple
> environments and I need to create group and multiple entries to switch
> environments.
> In this case I have implemented something complicated and I think not very
> well.
> 1) Ansible start my Zeppelin instance
> 2) Ansible perform a curl call to generate interpreter.json inside
> $ZEPPELIN_HOME/conf (this file seems to be lazy)
> 3) Ansible generate cassandra json with my configuration for environments
> with Jinja template
> 4) Ansible merge theses Jsons with $ZEPPELIN_HOME/conf/interpreter.json
>
> It's work but I suppose that it exists a better process to load Interpreter
> configurations.
>
> Thx for your help
>
> Baptiste
>
>
>
>
>
>
--
Best Regards
Jeff Zhang
