You will need to set at least the CPU count per node in slurm.conf and
we'll try to make this more clear in the logs and documentation. This
is from "man slurm.conf".
FastSchedule
0 Base scheduling decisions upon the actual configuration of
each individual node except that the node's processor count
in SLURM's configuration must match the actual hardware
configuration if SchedulerType=sched/gang or
Select‐
Type=select/cons_res are configured (both of those plugins
maintain resource allocation information using bitmaps for
the cores in the system and must remain static, while the
node's memory and disk space can be established later).
Quoting Tal Hazan <tha...@doc.com>:
> In our compute nodes we have redhat 6.2 with X5690 cpu's.
>
> -----Original Message-----
> From: Moe Jette [mailto:je...@schedmd.com]
> Sent: Wednesday, April 11, 2012 8:20 PM
> To: slurm-dev
> Subject: [slurm-dev] RE: CPU detection failure when using CR_Core
>
>
> Exactly what operating system and processor does your system have?
>
> Quoting Tal Hazan <tha...@doc.com>:
>
>> Hi,
>>
>> I tried version 2.3 and this issue occurs also in this version.
>>
>> Tal
>>
>> From: je...@schedmd.com [mailto:je...@schedmd.com]
>> Sent: Saturday, April 07, 2012 4:39 AM
>> To: slurm-dev
>> Subject: [slurm-dev] RE: CPU detection failure when using CR_Core
>>
>> try slurm v2.3. if that fails we will need to investigate why the code
>> fail with your distro and processor
>> --
>> Sent from my Android phone. Please excuse my brevity and typos.
>> Tal Hazan <tha...@doc.com<mailto:tha...@doc.com>> wrote:
>>
>> Hi Moe,
>>
>> I have FastSchedule set to 0
>> and
>> SelectType=select/cons_res
>> SelectTypeParameters=CR_Core
>> NodeName=node[001,002]
>>
>> [2012-04-07T03:31:14] slurmd version 2.2.7 started
>> [2012-04-07T03:31:14] slurmd started on Sat 07 Apr 2012 03:31:14 +0300
>> [2012-04-07T03:31:14] Procs=1 Sockets=1 Cores=1 Threads=1
>> Memory=96865 TmpDisk=1922 Uptime=34010
>>
>> Without slect/cons_res and CR_Core:
>>
>> [2012-04-07T03:32:59] slurmd version 2.2.7 started
>> [2012-04-07T03:32:59] slurmd started on Sat 07 Apr 2012 03:32:59 +0300
>> [2012-04-07T03:32:59] Procs=24 Sockets=2 Cores=6 Threads=2
>> Memory=96865 TmpDisk=1922 Uptime=34115
>>
>> Tal
>>
>> -----Original Message-----
>> From: Moe Jette [mailto:je...@schedmd.com]
>> Sent: Saturday, April 07, 2012 3:27 AM
>> To: slurm-dev; Tal Hazan
>> Subject: Re: [slurm-dev] CPU detection failure when using CR_Core
>>
>> You probably have FastSchedule=1 or 2.
>> Set to 0 if you want information loaded from the compute node.
>>
>> Quoting Tal Hazan <tha...@doc.com<mailto:tha...@doc.com>>:
>>
>>> Hi,
>>>
>>> I'm evaluating implementation of slurm and found an issues using it
>>> with CR_Cores type.
>>>
>>> What I'm are trying to achieve is assignment of cores per job
>>> (non-mpi) so we did this change in order to make Cores a consumable
>>> resource. And used #SBATCH -cores-per-socket=1 in batch jobs.
>>>
>>> Adding the following lines in slurm.conf causing the hardware
>>> detection mechanism to fail:
>>> SelectType=select/cons_res
>>> SelectTypeParameters=CR_Core
>>>
>>> And then I need to enter the resources manually I.E:
>>> NodeName=node[001,002] Procs=24 Sockets=2 CoresPerSocket=6
>>> ThreadsPerCore=2 RealMemory=96865 TmpDisk=1922
>>>
>>> Without adding it manually I see , single cpu, single socket, single
>>> thread and single core in each node.
>>>
>>>
>>> Thanks,
>>> Tal Hazan
>>>
>>>
>>
>>
>
>
