[torqueusers] qstat call in prologue or epilogue script

Garrick Staples garrick at usc.edu
Wed Sep 8 09:39:24 MDT 2004


On Wed, Sep 08, 2004 at 03:03:04PM +0200, cyrille martin alleged:
> Hi,
> 
> I would like know nodes and processors resources allocated for a job id 
> in prologue or epilogue script.
> 
> When i do a "qstat -f jobid" call in the prologue script the qstat hangs 
> and prologue is killed by the timeout as said in Administrator guide p43.
> When one node is allocated it works fine, but for more nodes the qstat 
> hangs.
> I use torque-1.1.0p0.
> 
> Is it a normal behavior or not, have you an idea?

It is normal.  pbs_server blocks during the prologue/epilogue scripts, so it is
unresponsive when you call qstat.

You can get the information from pbs_mom's directory.  The first argument to
prologue/epilogue is the jobid, then just look at $PBS_HOME/aux/$jobid

-- 
Garrick Staples, Linux/HPCC Administrator
University of Southern California
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20040908/06d9407b/attachment.bin


More information about the torqueusers mailing list