using zeppelin spark interpreter with the operator

[mvu-tk]

Hi team,

I'm trying to use zeppelin with the operator but I don't know how to use it.
Normally zeppelin interpreter would act as the driver program.
How can this fit into the operator world?

thanks,
-M

[jdonnelly-apixio]

I haven't tried setting up zeppelin, but something a little custom like this could probably work. I've had success with running the spark thrift server with this method. If you try this route and are successful, could you post details? At some point, I want to provide notebook functionality for my clusters and am contemplating using zeppelin with the spark thrift work around method as a stop gap before we get to jupyter notebooks provided by kubeflow (kubeflow uses spark on k8s operator for spark stuff, is pretty complex, but incredibly promising).