[torqueusers] Torque 2.5.x->4.x upgrade finding already-running jobs?

[Lloyd Brown]

Hi, all.

We're considering upgrading our production cluster from Torque 2.5.9 to
4.1.0, and wondered if I could ask a question, to tap into the
community's expertise.

I know that, due to communication-protocol changes, we need to upgrade
the pbs_server and pbs_mom's together.  But is there any known issue
with the 4.x pbs_mom's picking up on the already-running jobs (started
by the 2.5.x pbs_mom)?  Obviously the pbs_mom process won't be the
parent process of the job, but that shouldn't be any different than
restarting with the "-p" option anyway.

I'll do some testing too, but I just thought I'd ask around to see if
anyone has done it already, and what success/failures they've had.

If it works, it could save us most of a full-system outage, which is a
very big deal.
-- 
Lloyd Brown
Systems Administrator
Fulton Supercomputing Lab
Brigham Young University
http://marylou.byu.edu