[torqueusers] Enabling BLCR on Torque roll (problems with "qhold")

Andrus, Brian Contractor bdandrus at nps.edu
Tue Mar 15 13:06:40 MDT 2011


Ricardo,

I have started working on this too.
A couple things I have seen so far:

1) The sample blcr_checkpoint_script has some issues (it is missing a
comma on line 36 and $depth is not defined on line 21)

2) The # of arguments passed to the script by pbs_mom is 8. If the
script does not receive exactly 7, it is not happy.

3) What arguments are passed and in what order does not seem to be
documented and it _appears_ that the sample script is expecting
something other than what is being passed (eg: it is trying to 'cd' to
the headnode name)


So my question to the list:

Does anyone know the arguments and order that pbs_mom passes to the
checkpoint script?
Or
Does anyone have a working checkpoint script?


Brian Andrus
ITACS/Research Computing
Naval Postgraduate School
Monterey, California
voice: 831-656-6238

-----Original Message-----
From: torqueusers-bounces at supercluster.org
[mailto:torqueusers-bounces at supercluster.org] On Behalf Of Ricardo Alves
Sent: Tuesday, March 15, 2011 11:28 AM
To: torqueusers at supercluster.org
Subject: [torqueusers] Enabling BLCR on Torque roll (problems with
"qhold")

Hi,

Torque was compiled adding the "--enable-unixsockets=no --enable-blcr"
line to the ./configure command (as it says on the torque tutorial -
http://www.clusterresources.com/torquedocs21/2.6jobcheckpoint.shtml).

Both scripts are present in the mom_priv directory with execution
privileges and the paths are in the mom_priv/config (the config file is
attached to this mail)

thank you.


More information about the torqueusers mailing list