Hi all,
first of all thanks for this really nice package. We are using it in our HPC cluster and has been very good so far.
I just found that the pyslurm.db.tres.GPU.from_tres method raise an exception if the GPU tres type has more than one :. This is not very common but it should just be ignored and actually the fix is just one change to the tres.name.split line.
The snipped below reproduces the issue:
from pyslurm import db
if __name__ == "__main__":
jobs = db.Jobs().load()
And this is the error:
Traceback (most recent call last):
File "test.py", line 4, in <module>
jobs = db.Jobs().load()
File "pyslurm/db/job.pyx", line 271, in pyslurm.db.job.Jobs.load
File "pyslurm/db/tres.pyx", line 325, in pyslurm.db.tres.TrackableResources.load
File "pyslurm/db/tres.pyx", line 282, in pyslurm.db.tres.TrackableResources._handle_tres_type
File "pyslurm/db/tres.pyx", line 394, in pyslurm.db.tres.GPU.from_tres
ValueError: too many values to unpack (expected 2)
Of course, I'll be happy to create a PR with the fix.
Thanks!
Hi all,
first of all thanks for this really nice package. We are using it in our HPC cluster and has been very good so far.
I just found that the
pyslurm.db.tres.GPU.from_tresmethod raise an exception if the GPU tres type has more than one:. This is not very common but it should just be ignored and actually the fix is just one change to thetres.name.splitline.The snipped below reproduces the issue:
And this is the error:
Of course, I'll be happy to create a PR with the fix.
Thanks!