How to monitor idle kernels/jupyters

[Thierry Parmentelat]

Hello everybody

I am in the process of writing my own local frontend for serving notebooks to a MOOC infrastructure

my app creates jupyter dockers on the fly, but I also need to tear them down at some point

and as much as I know exactly when my users show up, I have no idea when they are actually done...

I guess a first implementation would be to give everybody a fixed duration after which the container gets trashed no matter what, but that's clearly a little rough, to say the least, and suboptimal

So I am curious to know if there are any mechanism at all that I can use to get a clue about when a user (or more specifically a jupyter inside a docker container) becomes idle in some sense, so that I could implement some kind of timeout based on that

thanks for any clue -- Thierry

[MinRK]

There are two ways I can think of right now:

1. use a traffic-monitoring proxy in front of the notebook server, such as configurable-http-proxy and monitor activity there.
2. Wait for notebook 5.0 (hopefully soon), which will have activity tracking in the API.

-Min

​

--
You received this message because you are subscribed to the Google Groups "Project Jupyter" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jupyter+unsubscribe@googlegroups.com.
To post to this group, send email to jup...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jupyter/d904f3cb-98a4-4248-bd85-3d278becdd32%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.