[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] Submitting to condor via globus

Steven C. Timm, Ph.D  (630) 840-8525  timm@xxxxxxxx  http://home.fnal.gov/~timm/
Fermilab Computing Div/Core Support Services Dept./Scientific Computing Section
Assistant Group Leader, Farms and Clustered Systems Group
Lead of Computing Farms Team

On Thu, 6 Jul 2006, Greg.Hitchen@xxxxxxxx wrote:

Just a quick question to see if I understand the logic of allowing jobs
to be submitted to a
condor pool/s via Globus.

Linux box with both globus and condor installed on it.

Job manager queues setup on globus for each arch/opsys in the condor

Job submitted via globus for the condor pool. I assume no issues with
submitting to
windows machines in the condor pool? I also assume that flocking is OK?

Never heard of anyone who has done a globus submit to windows machines.

Flocking is OK (for Globus GT2) only if all the machines in the pools
which jobs flock to have a shared file system with the machine
which runs the globus gatekeeper and the GRAM interface.  Globus GT2
depends on NFS to share the proxy with all the compute nodes and also
the executable, input, and standard out/err.  With condor-G there are some
ways to work around some of that but you still need the NFS.

So the globus job end up being submitted to the condor schedd on the
same box?

Yes that is the default.  Quite tricky to configure it otherwise.




P.S. I'll worry about details later! :)

Greg Hitchen
CSIRO Exploration and Mining				phone:+61 8 6436
Australian Resources Research Centre (ARRC)	fax:	+61 8 6436 8555
Postal address:						mob:	0407 952
PO Box 1130, Bentley WA 6102, Australia
Street Address:
26 Dick Perry Avenue, Kensington WA 6151