They are linked because you can have multiple interpreters with the same
type.

For example a spark interpreter using a different cluster, or with less
core or memory
On Jun 17, 2015 7:56 PM, "Ophir Cohen" <[email protected]> wrote:

> Actually I encounter a problems with just copying the notebook dir as
> there is connection to the interpreter.json file (if a notebook isn't in
> the interpreter file you'll be asked all the time for the interpreters).
> * Zeppelin guys: why we need the link from interpreters to notebooks, why
> not assume that ALL notebooks use ALL interpreters?
> IMHO it just open the option to inconsistencies.
>
>
> I've ended up creating scripts that taking care of the data consistency:
> 1. Check which notebook exists and add them to the interpreter.json file.
> 2. Check which notebook does not exist and remove.
>
> Ophir
>
> PS
> I also developed a syncing mechnism for the notebooks in order to share it
> across clusters if somebody intersting I can provide some details.
>
> On Sun, Jun 14, 2015 at 5:01 PM, Silvio Fiorito <
> [email protected]> wrote:
>
>>  There no official way, but you can simply copy the folder into the
>> notebook path and restart Zeppelin.
>>
>>  *From:* Ramani, Ashok <[email protected]>
>> *Sent:* ‎Wednesday‎, ‎June‎ ‎10‎, ‎2015 ‎12‎:‎22‎ ‎PM
>> *To:* [email protected]
>>
>>  Hi,
>>
>>  Is there a way to import a notebook into the Zeppelin UI? Or, start a
>> notebook from a template?
>>
>>  Thanks,
>> Ashok
>>
>>
>

Reply via email to