-
Notifications
You must be signed in to change notification settings - Fork 119
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Provide a way to access the GRES_IDX attribute for a job #116
Comments
Thanks. I was hoping this one would be easy. Here is the code that needs to get wrapped: https://github.com/SchedMD/slurm/blob/5073024350eb79c8c5a9964e800bc0ce3ab93d59/src/api/job_info.c#L706-L803. That may take me some time. How would you like the output of this attribute, as one string that matches the scontrol output, or a list of strings if there are more than one? |
Sorry for not spotting this before, I've just passed your query on to the developer! |
Got a reply from the developer today, he said:
Hope that helps! |
Hi @chrissamuel. I know it's been a while, but I think I got the code wrapped to get |
Hi @giovtorres! Swinburne is on 18.08.x now, but I've since moved to the US for love and for work and am now at NERSC (still doing HPC). But I'll still see updates here and let them know. Thanks for this! |
The |
Thanks! Passed that back to them. |
Details
Issue
It would be really useful to us for our job monitoring program to be able to get the
GRES_IDX
information from running jobs so we can highlight the GPU the job is using for our users so they can see how much utilisation it is getting.Our issue for this is: conradtchan/jobmon#1
This looks like
GRES_IDX=gpu(IDX:0)
orGRES_IDX=gpu(IDX:1)
orGRES_IDX=gpu(IDX:0-1)
for our systems with dual GPUs. I'm not sure what it would look like for a system with say 4 GPUs if the allocation is not contiguous.Thanks for considering this!
All the best,
Chris
The text was updated successfully, but these errors were encountered: