How to increase number of CPU/ Nodes in Singularity run

80 views
Skip to first unread message

Nidhal Ghanmi

unread,
Nov 30, 2021, 1:41:16 AM11/30/21
to singularity
Hello

excuse me for maybe basic questions
I am new in singularity and I did run an image , but it looks like using only one cpu

Any way to increase it ?
singularity version 3.6.3
and alos I am using it on an HPC , I did not yet run it with the scheduler , but if there is alos additional configurations please let me know also to run it with multi nodes


Thanks
Nidhal

Thomas Hartmann

unread,
Nov 30, 2021, 5:52:27 AM11/30/21
to singu...@lbl.gov
Hi Nidhal,

normally, Singularity does not handle resource management, i.e., the cap
to one core is originating from somewhere else.

Have you checked, that your HPC jobs are requesting sufficient cores and
get them allocated?

Cheers,
Thomas
> --
> You received this message because you are subscribed to the Google
> Groups "singularity" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to singularity...@lbl.gov
> <mailto:singularity...@lbl.gov>.
> To view this discussion on the web visit
> https://groups.google.com/a/lbl.gov/d/msgid/singularity/9245bfd7-2ee8-4dc1-8430-a4330a828c64n%40lbl.gov
> <https://groups.google.com/a/lbl.gov/d/msgid/singularity/9245bfd7-2ee8-4dc1-8430-a4330a828c64n%40lbl.gov?utm_medium=email&utm_source=footer>.

Joseph Schoonover

unread,
Nov 30, 2021, 9:43:24 AM11/30/21
to singu...@lbl.gov
Hey Nidhal,
Is your application build with MPI, OpenMP, or some other framework that allows you to run on multiple cores?

To unsubscribe from this group and stop receiving emails from it, send an email to singularity...@lbl.gov.
To view this discussion on the web visit https://groups.google.com/a/lbl.gov/d/msgid/singularity/1702d6c1-bb88-3bf6-d591-aebbfc8218f5%40desy.de.

Nidhal Ghanmi

unread,
Nov 30, 2021, 10:05:16 AM11/30/21
to singularity, Joseph Schoonover
Hello

Thanks Thomas and Joseph

I am willing to use it with openmpi but not with that test
Also , there was running jobs on all nodes , so maybe that is the reason ? I will retry once HPC nodes are free

and I am trying to learn more about this mpi also , it s new for me

Thanks
Nidhal
Reply all
Reply to author
Forward
0 new messages