Guillaume Eynard-Bontemps
Guillaume Eynard-Bontemps
Hi everyone, very interested in this too, as I am currently deploying a jupyterhub instance on our PBS HPC Cluster. Thanks @mrocklin for pointing that to me. @rkdarst, your outline...
Hi, Is there any advance here? I tried to play with the wiki but, it seems I'm not able to do any modifications. I've deployed a beta version of Jupyterhub...
I think that implementing #133 and give the ability to override the template would be a good solution.
> One approach would be to allow users to explicitly skip lines that their cluster manager finds troublesome. I put this into a commit in the spec-rewrite branch for now...
Yes, that's the idea. Using Jinja2 should allow conditionnals!
> Nothing. I suspect that it is defaulting to giving me one full node per job 😄 Yeah, with Summit, there's no point using less than a full node!!
Closing this: `header_skip` kwarg is implemented, for the rest we should rely on a Jinja template.
With PBS Pro, you've got the `qpeek $JOB_ID` command to get the stdout, and `qpeek -e $JOB_ID` for stderr. As far as I know, output files are not availabe until...
Here is how it looks: 
I guess I didn't know what to do with @mrocklin comments... I feel this is applicable to other cluster managers, but I don't know how to standardize it, so that...