wrong gres_length

[back2net]

Hi, in our cluster, we use many devices with long names, and system-status dashboard does'not count number of Active GPUs right.
So, in ood_core/lib/ood_core/job/adapters/slurm.rb:113, I was changed correction value from 2 to 100, and it is work's. I'am not ruby dev, so i can't make right fix.

gres_length = call("sinfo", "-o %G").lines.map(&:strip).map(&:length).max + 2 <= here

OnDemand version: 4.0.1

[johrstrom]

🤦‍♂ we should probably have just set it to 100 and chomped all the whitespace. That way we don't make an extra call to sinfo.