* Ole Holm Nielsen <ole.h.niel...@fysik.dtu.dk> [190903 11:14]: > How do you dynamically update your gres=localtmp resource according to the > current disk free space? I mean, there is already a TmpFS disk space size > defined in slurm.conf, so how does your gres=localtmp differ from TmpFS?
Dear Ole, I think (but please correct me, if I'm wrong) the `TmpDisk=xxxxxxxxx´ disk space size defined in slurm.conf is just a fixed node property. Thus, specifying e.g. `--tmp=1000000´ does prevent a job from allocating a node that has less than 1000000 MB disk size configured in slurm.conf but does not count against the amount of disk space that is left available for other jobs on that nodes. In other words: TmpDisk does not seem to work like a consumable resource and `--tmp=1000000´ does not seem to behave like a consumable resource allocation, whereas a GRES based specification for local scratch space does. At least that's how I understood the documentation. Did I get this wrong? Best regards Jürgen -- Jürgen Salk Scientific Software & Compute Services (SSCS) Kommunikations- und Informationszentrum (kiz) Universität Ulm Telefon: +49 (0)731 50-22478 Telefax: +49 (0)731 50-22471
