RStudio from Jupyterhub

I have a Jupyterhub installation that spawns notebooks [ie, Docker Images] into a Kubernetes cluster (or which a Jupyter Notebook with R 3.5 is one)

I've been asked to provide a notebook with R-Studio - and I've [mostly] done that (I can provide the Dockerfile if people are interested).... and by mostly done, I mean that the demo() examples run

I've one, maybe two, issues to resolve - for which I need some help:

  1. Once the user quits the R-Studio session, the only way to restart is to restart the whole docker image... is there a way to restart an R-Studio session without doing that?

  2. Is there a way to get the Jupyter Notebook to auto-start the R-Studio session, and not have to select from the new menu once the docker image has started?

Those 2 are limitations of the current implementation of jupyter-rsession-proxy and are kinda related.

There are two options:

  1. Fixing/adding a way for the jupyter-proxy-server to auto-start and then changing the default path of the application to be something like rstudio: https://github.com/jupyterhub/jupyter-server-proxy

  2. You could also have another image that just starts rstudio and doesn't use the jupyter-server-proxy, that way uses can pick what image they want to start, one for jupyter and one for rstudio. In that case you will also change the config variable above.

Those are not perfect solutions but maybe help a bit.

It is possible that some one here uses a similar setup, but I'm afraid this is not the best place for asking this kind of questions (RStudio doesn't officially support jupyter products). Maybe you would have better luck asking this on the jupyterhub community.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.