[lug] Processor assignment

Davide Del Vento davide.del.vento at gmail.com
Wed Mar 30 21:42:37 MDT 2016


Is this a single server or multiple servers? If it's a single one, why
are you using MPI in the first place? It's overkill. If it's multiple
server, why aren't you running a resource manager such as slurm?

On Wed, Mar 30, 2016 at 4:20 PM, Rob Nagler <nagler at bivio.biz> wrote:
> On Wed, Mar 30, 2016 at 4:11 PM, Zan Lynx <zlynx at acm.org> wrote:
>>
>> To be sure, you aren't doing anything with resource limits, are you? This
>> link includes the things you can do. And if you're running Kubernetes, these
>> will happen automatically.
>>
>> https://docs.docker.com/engine/reference/run/#runtime-constraints-on-resources
>
>
> Thanks, and sorry for the confusion. This has been reproduced natively on
> CentOS 6 and Fedora 23. It has not been reproduced natively on Debian
> Jessie, but it has been reproduced on the same machine inside a single
> container. This last fact *may* be a clue or a red herring.
>
> FYI, Kubernetes and MPI don't play nicely together. :-(
>
> Rob
>
>
> _______________________________________________
> Web Page:  http://lug.boulder.co.us
> Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug
> Join us on IRC: irc.hackingsociety.org port=6667 channel=#hackingsociety


More information about the LUG mailing list