[Mauiusers] requesting cores with ncpus

Mahmood Naderan nt_mahmood at yahoo.com
Sun Jan 9 23:46:28 MST 2011


>Use nodes=18 instead of ncpus.W/o ppn nodes=X means requesting X number of CPU's 
>w/o any geametry specification

When I chnaged the nodes=X (X>2), I get an error that there are no such 
resources. Then I found a *magic* queue command in 
http://www.clusterresources.com/pipermail/torqueusers/2006-January/002869.html that
 did the job. Now when I request nodes=18, it allocates 16 cores from 
one station and 2 cores from the other station. 


That command is:
qmgr -c "set server resources_available.nodect=32" 
where you can replace 32 with number of total cores you have in your cluster.

Thanks, 
// Naderan *Mahmood;




________________________________
From: rishi pathak <mailmaverick666 at gmail.com>
To: Mahmood Naderan <nt_mahmood at yahoo.com>
Cc: mauiusers at supercluster.org
Sent: Mon, January 10, 2011 9:05:23 AM
Subject: Re: [Mauiusers] requesting cores with ncpus

Hi Mahmood,
                         Use nodes=18 instead of ncpus.W/o ppn nodes=X means 
requesting X number of CPU's w/o any geametry specification


On Sun, Jan 9, 2011 at 8:40 PM, Mahmood Naderan <nt_mahmood at yahoo.com> wrote:

Hi,
>Recently I submitted a message to torque user group but found that my question
>is actually related to maui. Here is the problem:
>With two stations (server+worker) each has 16 cores, I want to submit a job but
>don't want to explicitly mention on which station. 
>If I use "nodes=2:ppn=4", then 4 cores on each station is devoted. That is
>fine...
>If I use "ncpus=18", I expect that maui put 16 processes on one station and 2 
on
>the other automatically. However the actual result is that my job is deferred.
>Total Tasks: 1  Req[0]  TaskCount: 1  Partition: ALL Network: [NONE]  Memory >=
>0  Disk >= 0  Swap >= 0 Opsys: [NONE]  Arch: [NONE]  Features: [NONE] Dedicated
>Resources Per Task: PROCS: 17 NodeCount: 1   IWD: [NONE]  Executable:  [NONE]
>Bypass: 0  StartCount: 0 PartitionMask: [ALL] Flags:       RESTARTABLE  job is
>deferred.  Reason:  NoResources  (cannot create reservation for job '152'
>(intital reservation attempt) ) Holds:    Defer  (hold reason:  NoResources) 
PE:
>17.00  StartPriority:  1 cannot select job 152 for partition DEFAULT (job hold
>active)
>
> 
> 
>How can I fix that?
>Thanks, 
>// Naderan *Mahmood;
>
>
>
>_______________________________________________
>mauiusers mailing list
>mauiusers at supercluster.org
>http://www.supercluster.org/mailman/listinfo/mauiusers
>


-- 
Regards--
Rishi Pathak
National PARAM Supercomputing Facility
Center for Development of Advanced Computing(C-DAC)
Pune University Campus,Ganesh Khind Road
Pune-Maharastra



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/mauiusers/attachments/20110109/c087ecd8/attachment.html 


More information about the mauiusers mailing list