[slurm-users] [slurm-us> , -- , Ole Holm Nielsen, PhD, Senior HPC Officer, Department of Physics, Technical University of Denmark, , Building 307, DK-2800 Kongens Lyngby, Denmark, E-mail: Ole.H.Nielsen at fysik.dtu.dk , Homepage: http://dcwww.fysik.dtu.dk/~ohnielse/, Tel: (+45) 4525 3187 / Mobile (+45) 5180 1620 ers] Having Issue in Slurm cluster setup
Ole Holm Nielsen
Ole.H.Nielsen at fysik.dtu.dk
Tue Apr 9 10:14:42 UTC 2019
Your slurm.conf line doesn't specify the node's physical memory:
NodeName=ozd2485u Gres=gpu:2 Sockets=2 CoresPerSocket=14
ThreadsPerCore=2 State=UNKNOWN
See "man slurm.conf":
RealMemory
Size of real memory on the node in megabytes (e.g.
"2048"). The default value is 1.
On 4/9/19 8:47 AM, sudhagar s wrote:
> Attaching my slurm.conf file. can you please help me to find the issue.
>
> On Tue, Apr 9, 2019 at 12:08 PM Ole Holm Nielsen
> <Ole.H.Nielsen at fysik.dtu.dk <mailto:Ole.H.Nielsen at fysik.dtu.dk>> wrote:
>
> On 09-04-2019 08:33, sudhagar s wrote:
> > Thanks Ole,
> >
> > when i give "scontrol show node" it list down the details. where
> i can
> > see RealMemory=1 is this will be a problem?
>
> In your "scontrol show node" image I read RealMemory=1 (units of MB)
> and
> mem=1M. I think you configured slurm.conf incorrectly.
>
> > On Tue, Apr 9, 2019 at 11:53 AM Ole Holm Nielsen
> > <Ole.H.Nielsen at fysik.dtu.dk <mailto:Ole.H.Nielsen at fysik.dtu.dk>
> <mailto:Ole.H.Nielsen at fysik.dtu.dk
> <mailto:Ole.H.Nielsen at fysik.dtu.dk>>> wrote:
> >
> > On 09-04-2019 07:37, sudhagar s wrote:
> > > Hi, Iam newbee in slurm. trying to setup a cluster for ML
> training
> > > purpose. i created controle node and compute node. both are up
> > and running.
> > >
> > > when i enter "srun -N 1 hostname" it says
> > > " srun error memory specification can not be satisfied"
> > > "unable to allocate resources: requested node
> configuration is not
> > > available"
> > >
> > > how to fix this?
> >
> > Probably you made some errors in configuring slurm.conf.
> Look at your
> > NodeName and PartitionName definitions to figure out why the
> resources
> > are incorrect.
> >
> > /Ole
More information about the slurm-users
mailing list