dingzeyuli / seclumon

Server Cluster Monitor (SeCluMon). A python/cgi based monitoring system I built for our graphics group cluster.

Hi there 👋

🏡    I am a Senior Research Scientist at Adobe Research.

🎓    I create new interactions for video creation and authoring workflow. I am a heavy user of deep learning, LLMs, and GenAI.

🔎    I’m always looking for interns and collaborators to work on research projects leading to publications/patents/product features. Please feel free to reach out to learn more.

📫    How to reach me -- see http://dingzeyu.li/.

seclumon's Issues

show the color box according to the worst component

add disk storage display

can we add scheduler on top of this monitor?

Frustrated by running lots of server jobs, @odedstein suggests to add simple scheduler functionalities to SeCluMon. I think it might be possible.

• In the backend, we will store a list of to-be-executed commands.
• Whenever refresh.py is run, it not only update the cpu/ram/temperature display info, but also put new jobs on the idle clusters.
• this job allocation might be done through a nohup ssh session. Oded might know how to do this.

Electron app?

Suggested by @breannansmith.

Will look into it later.

change the color display to box-wise

to see each machine is bounded by CPU or RAM or disk space or temperature

suggested by Breannan