[torqueusers] time conflict

[Garrick Staples]

On Wed, Oct 11, 2006 at 12:16:33PM -0300, Albino Aveleda alleged:
> Hi,
> 
> I have a cluster with 32 nodes with one CPU per node.
> 
> All queue is working, except the dedicated queue. I don't know, why? 
> When I submited one job to use 32 nodes (with "#PBS -l nodes=32").
> It always still in queued (status Q). When I digit "qstat -f <jobid>" 
> the coment is:
> 
> coment = Not Running: Dedicated time conflict
> 
> I have sincronized the date and time among nodes and server with ntp.
> What is wrong?
> 
> Summary of torque configuration is below and I've put the queues for 8 
> cpus and dedicated.

Change this line in your sched_config:

# all queues starting with this value are dedicated time queues
# i.e. dedtime or dedicatedtime would be dedtime queues
#       NO PRIME OPTION
dedicated_prefix: ded