Memory bound problem with Docker

[Jimmy Liu]

Dear all,

I built a Docker image from dealii/dealii:v9.2.0-focal from the dockerhub. When I ran mpirun I got the following warning

WARNING: Open MPI tried to bind a process but failed.  This is
warning only; your job will continue, though performance may
be degraded.

  Local host:        8f3a6d698b99
  Application name:  ./myprogram
  Error message:     failed to bind memory
  Location:          rtc_hwloc.c:447

I also saw the similar warning as the program ran

3 more processes have sent help message help-orte-odls-default.txt / memory not bound
Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

I can still  gain some speed using multiple cores, but the gain is not that much.

Neither the --bind-to core or hwthread option can solve this issue.

Do you have any thoughts? Thank you!

Best Regards,

Jimmy

[Wolfgang Bangerth]

Jimmy:
I have no experience with this, but assume that you just can't tie programs to
processors when running under your docker installation. That's what I see
here, for example:
https://github.com/open-mpi/ompi/issues/7368
Best
W>

On 2/14/24 15:21, Jimmy Liu wrote:
> *** Caution: EXTERNAL Sender ***

>
> Dear all,
>
> I built a Docker image from dealii/dealii:v9.2.0-focal from the dockerhub.
> When I ran mpirun I got the following warning
>

> *WARNING: Open MPI tried to bind a process but failed.  This is

> warning only; your job will continue, though performance may
> be degraded.
>
>   Local host:        8f3a6d698b99
>   Application name:  ./myprogram
>   Error message:     failed to bind memory

>   Location:          rtc_hwloc.c:447*

>
> I also saw the similar warning as the program ran
>

> *3 more processes have sent help message help-orte-odls-default.txt / memory

> not bound
> Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error

> messages*

>
> I can still  gain some speed using multiple cores, but the gain is not that much.
>
> Neither the --bind-to core or hwthread option can solve this issue.
>
> Do you have any thoughts? Thank you!
>
> Best Regards,
> Jimmy
>

> --
> The deal.II project is located at http://www.dealii.org/
> <http://www.dealii.org/>
> For mailing list/forum options, see
> https://groups.google.com/d/forum/dealii?hl=en
> <https://groups.google.com/d/forum/dealii?hl=en>
> ---
> You received this message because you are subscribed to the Google Groups
> "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to dealii+un...@googlegroups.com
> <mailto:dealii+un...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/dealii/587120e0-cd23-48eb-95fb-6da6b5eaef12n%40googlegroups.com <https://groups.google.com/d/msgid/dealii/587120e0-cd23-48eb-95fb-6da6b5eaef12n%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/