limit the number of simultaneous servers

[rokroskar]

When a user requests a new server, check how many are already running and if that number is equal to MAX_USER_SERVERS then shut down the oldest one before starting the new one.

Note (@leafty): shutdown oldest from the same (project, user). The idea is that the user is iterating with his/her Docker image

[lorenzo-cavazzi]

this should be possible starting from version 1.0 by using c.JupyterHub.named_server_limit_per_user https://jupyterhub.readthedocs.io/en/stable/changelog.html#new-features

[rokroskar]

We probably want to resolve #222 and SwissDataScienceCenter/renku-ui#686 before imposing these limits.

[olevski]

Should this be moved to Amalthea?

[rokroskar]

Not sure I think it should be the notebook service concern since it's also the one initiating new sessions. Why would it go to amalthea? It doesn't really know the concept of a user afaik.

[olevski]

Yeah that is a good point @rokroskar I did not read the description well enough. I thought this talked about limiting the total number of all servers on a cluster. Not the number per user.