Dear all,
I did a little bit more testing.
* I have reenabled CR_ONE_TASK_PER_CORE.
* My testnode is still configured, as slurmd -C tells me.
* "--ntasks=24" or "--ntasks=24 --ntasks-per-node=24" can both be
submitted, resulting in a job with the "free" hyperthread per task.
Nearly perfect.
BUT:
The node has 48 CPUs:
NodeName=lnm596 Arch=x86_64 CoresPerSocket=12
CPUAlloc=0 CPUTot=48 CPULoad=0.04
but I cannnot submit the following:
sbatch -p test -n 24 --ntasks-per-node=24 --cpus-per-task=2 -w lnm596
24*2 is 48, so I'm asking for 48 CPUs.
There is still something wrong with CR_ONE_TASK_PER_CORE.
Best
Marcus
On 2/20/19 7:49 AM, Chris Samuel wrote:
On Tuesday, 19 February 2019 10:14:21 PM PST Marcus Wagner wrote:
sbatch -N 1 --ntasks-per-node=48 --wrap hostname
submission denied, got jobid 199805
On one of our 40 core nodes with 2 hyperthreads:
$ srun -C gpu -N 1 --ntasks-per-node=80 hostname | uniq -c
80 nodename02
The spec is:
CPUs=80 Boards=1 SocketsPerBoard=2 CoresPerSocket=20 ThreadsPerCore=2
Hope this helps!
All the best,
Chris
--
Marcus Wagner, Dipl.-Inf.
IT Center
Abteilung: Systeme und Betrieb
RWTH Aachen University
Seffenter Weg 23
52074 Aachen
Tel: +49 241 80-24383
Fax: +49 241 80-624383
wag...@itc.rwth-aachen.de
www.itc.rwth-aachen.de
