search

FAIRiCUBE / FAIRiCUBE-Hub-issue-tracker

FAIRiCUBE HUB issue tracker
Creative Commons Zero v1.0 Universal
0 stars 1 forks source link

Kernel usage details are not available in fairicube hub Jupyter - fairicubeuc2-torch kernel #28

Closed robknapen closed 4 weeks ago

robknapen commented 7 months ago

The fairicubeuc2-torch kernel (and perhaps others as well) might need some updating so that the kernel usage details can be displayed:

Screenshot 2023-11-13 at 10 26 55

eox-cs1 commented 4 weeks ago

we show machine resources via another installed tooling so no further action planned image

KathiSchleidt commented 4 weeks ago

@eox-cs1 so no integration with FAIRiCUBE requirements foreseen?

eox-cs1 commented 4 weeks ago

@KathiSchleidt What? I said that there is already a tool available to monitor used machine resources. So no need for another one.

jetschny commented 4 weeks ago

I could not test the resource estimate creation on EOX Hub, it is on my list (or to poke Bachir to do it), but this the kernel usage is an "add-on" to the "measurerer" by Antonio, not a conflicting solution. Rob, how do you use the kernel usage? if the kernel usage is used for monitoring but available outside the jupyter lab, it might be fine?

robknapen commented 4 weeks ago

It is this extension: https://github.com/jupyter-server/jupyter-resource-usage

Which allows a live view of the resource usage of the running notebook. I find it handy to keep an eye on long running / data intensive processes that are executed remotely. So a different purpose than the "measurerer". But it overlaps with the screenshot that @eox-cs1 included before, so should be covered.

jetschny commented 4 weeks ago

covered enough to close this issue?

robknapen commented 4 weeks ago

By the looks of it. So let me close the issue for now :)

eox-cs1 commented 4 weeks ago

it is available inside JupyterLab, va the sidebar