Prentice (and others) — if the NodeWeight/topology plugin interaction bothers
you, feel free to tack onto bug 6384.
https://bugs.schedmd.com/show_bug.cgi?id=6384
> On Jan 22, 2019, at 1:15 PM, Prentice Bisbal wrote:
>
> Killian,
>
> Thanks for the input. Unfortunately, all of this information
Killian,
Thanks for the input. Unfortunately, all of this information from you,
Ryan and others, is really ruining my plans, since it makes it look like
my plan to fix a problem wit my cluster will not be as easy to fix as
I'd hoped. One of the issues with my "Frankencluster" is that I'd like
Ryan,
Thanks for looking into this. I hadn't had a chance to revisit the
documentation since posing my question. Thanks for doing that for me.
Prentice Bisbal
Lead Software Engineer
Princeton Plasma Physics Laboratory
http://www.pppl.gov
On 1/18/19 2:58 PM, Ryan Novosielski wrote:
The docume
The documentation indicates you need it everywhere:
https://slurm.schedmd.com/topology.conf.html
"Changes to the configuration file take effect upon restart of Slurm daemons,
daemon receipt of the SIGHUP signal, or execution of the command "scontrol
reconfigure" unless otherwise noted."
I have
> On Jan 18, 2019, at 11:53 AM, Kilian Cavalotti
> wrote:
>
> On Fri, Jan 18, 2019 at 6:31 AM Prentice Bisbal wrote:
>>> Note that if you care about node weights (eg. NodeName=whatever001
>>> Weight=2, etc. in slurm.conf), using the topology function will disable it.
>>> I believe I was pro
On Fri, Jan 18, 2019 at 6:31 AM Prentice Bisbal wrote:
> > Note that if you care about node weights (eg. NodeName=whatever001
> > Weight=2, etc. in slurm.conf), using the topology function will disable it.
> > I believe I was promised a warning about that in the future in a
> > conversation wit
On 01/17/2019 07:55 PM, Fulcomer, Samuel wrote:
We use topology.conf to segregate architectures (Sandy->Skylake), and
also to isolate individual nodes with 1Gb/s Ethernet rather than IB
(older GPU nodes with deprecated IB cards). In the latter case,
topology.conf had a switch entry for each no
On 01/17/2019 06:36 PM, Ryan Novosielski wrote:
I don’t actually know the answer to this one, but we have it provisioned to all
nodes.
Note that if you care about node weights (eg. NodeName=whatever001 Weight=2,
etc. in slurm.conf), using the topology function will disable it. I believe I
ent:* Thursday, January 17, 2019 5:58 PM
> *To:* Slurm User Community List
> *Subject:* Re: [slurm-users] Topology configuration questions:
>
> We use topology.conf to segregate architectures (Sandy->Skylake), and also
> to isolate individual nodes with 1Gb/s Ethernet rather than IB (o
odes using the --constraint=whatever flag.
Nicholas McCollum
Alabama Supercomputer Authority
From: "Fulcomer, Samuel"
Sent: Thursday, January 17, 2019 5:58 PM
To: Slurm User Community List
Subject: Re: [slurm-users] Topology configuration questions:
We
We use topology.conf to segregate architectures (Sandy->Skylake), and also
to isolate individual nodes with 1Gb/s Ethernet rather than IB (older GPU
nodes with deprecated IB cards). In the latter case, topology.conf had a
switch entry for each node.
It used to be the case that SLURM was unhappy wi
I don’t actually know the answer to this one, but we have it provisioned to all
nodes.
Note that if you care about node weights (eg. NodeName=whatever001 Weight=2,
etc. in slurm.conf), using the topology function will disable it. I believe I
was promised a warning about that in the future in a
> On Jan 17, 2019, at 4:49 PM, Prentice Bisbal wrote:
>
> From https://slurm.schedmd.com/topology.html:
>
>> Note that compute nodes on switches that lack a common parent switch can be
>> used, but no job will span leaf switches without a common parent (unless the
>> TopologyParam=TopoOptional
And a follow-up question: Does topology.conf need to be on all the
nodes, or just the slurm controller? It's not clear from that web page.
I would assume only the controller needs it.
Prentice
On 1/17/19 4:49 PM, Prentice Bisbal wrote:
From https://slurm.schedmd.com/topology.html:
Note that
From https://slurm.schedmd.com/topology.html:
Note that compute nodes on switches that lack a common parent switch
can be used, but no job will span leaf switches without a common
parent (unless the TopologyParam=TopoOptional option is used). For
example, it is legal to remove the line "Switch
15 matches
Mail list logo
