[torqueusers] Creating a queue for certain nodes

シピオニ ロベルト SCIPIONI.Roberto at nims.go.jp
Tue Oct 21 16:53:17 MDT 2008


I tried the  option  

set queue node11 resources_default.neednodes =  node11

as well, with no avail  !!!

To be honest I think this is a limit of Torque .

Which routines should be corrected to improve the situation ?

RS




> I think you want
> 
> set queue node11 resources_default.neednodes =  node11
> 
>  instead of:
> 
> set queue node11 resources_default.nodes = node11
> 
> Hope this helps,
> 
> -Steve
> 
> 
> On Oct 20, 2008, at 9:54 PM, SCIPIONI Roberto wrote:
> 
> > Dear all,
> >
> >
> > I tried to create a queue that puts the jobs only in the node 11
> >
> > The extract from the command pbsnodes tells me that
> >
> > slavenode11
> >     state = free
> >     np = 8
> >     properties = big,8core,node11
> >     ntype = cluster
> >     status = opsys=linux,uname=Linux SLAVENODE11 2.6.5-7.244-smp 
> #1  
> > SMP Mon Dec 12 18:32:25 UTC 2005 x86_64,sessions=? 
> 15201,nsessions=?  
> > 15201 
> > ,nusers 
> > = 
> > 0 
> > ,idletime 
> > = 
> > 7342331 
> > ,totmem 
> > = 
> > 12366300kb 
> > ,availmem 
> > = 
> > 12271880kb 
> > ,physmem 
> > = 
> > 8157292kb 
> > ,ncpus=8,loadave=0.11,netload=64966591093066,state=free,jobs=?  
> > 15201,rectime=1224551444
> >
> >
> > so, it has been given a property     node11
> >
> > then I created a queue
> >
> >
> > create queue node11
> > set queue node11 queue_type = Execution
> > set queue node11 resources_default.nodes = node11
> > set queue node11 resources_default.walltime = 24:00:00
> > set queue node11 max_user_run = 5
> > set queue node11 enabled = True
> > set queue node11 started = True
> >
> >
> > But when I submit a job with the script
> >
> >
> > #!/bin/sh
> > ###########################################################
> > #PBS -N TEST
> > #PBS -l walltime=100:00:00
> > #PBS -l nodes=1:ppn=4
> > ############################################################
> > cd ${PBS_O_WORKDIR}
> > .
> > .
> >
> >
> > using the command
> >
> >
> > qsub -q node11 script.sh
> >
> >
> > the job would go to the fiorst available node on the list
> >
> > /server_priv/nodes
> >
> > and NOT on the node11.
> >
> > That differs to waht specified in the manual of Torque on 
> section  
> > 4.1.4
> >
> >
> > Any suggestions to why it does not work ?
> >
> >
> > Thanks
> >
> >
> >
> > Roberto Scipioni
> > WPI-MANA
> > Japan
> > _______________________________________________
> > torqueusers mailing list
> > torqueusers at supercluster.org
> > http://www.supercluster.org/mailman/listinfo/torqueusers
> 
> 


More information about the torqueusers mailing list