Dear friends,
my today topic is TruCluster and batch services (like NQS or PBS).
I'm running PBS
http://www.pbspro.com (which is very much like NQS
http://www.gnqs.org) for providing my users with batch queues for
running their CPU-bound numericals models.
The batch queues are workinkg on both an AS4100 and a ES40 which are
completely separate machines. Now I plan to merge the two machines in a
single TrueCluster 5.0a (or later). The question is:
how can I be sure that a running model is NOT given to any 4100
cpu while there are free ES40 cpus? My users would'nt like their
jobs to spend time on the slower machine while the faster is
available.
I'm quite sure I can force PBS to use processor_sets (4), but I
did'nt find info about psets existence in TruClusters. If psets
exists, I could create one of them containing all the ES40 cpus,
and then use pset_assign_pid to force submitted jobs to use that
pset.
I would like to receive comments and suggestions of any kind about the
managemente of batch queues in TruCluster environment.
Thanks a lot from Italy,
Emanuele
--
$$$ Emanuele Lombardi
$$$ mail: AMB-GEM-CLIM ENEA Casaccia
$$$ I-00060 S.M. di Galeria (RM) ITALY
$$$ mailto:emanuele.lombardi_at_casaccia.enea.it
$$$ tel +39 06 30483366 fax +39 06 30483591
$$$
$$$ |||
$$$ \|/ ;_;
$$$ What does a process need | /"\
$$$ to become a daemon ? | \v/
$$$ | |
$$$ - a fork o---/!\---
$$$ | |_|
$$$ | _/ \_
$$$* Contrary to popular belief, UNIX is user friendly.
$$$ It's just very particular about who it makes friends with.
$$$* Computers are not intelligent, but they think they are.
$$$* True programmers never die, they just branch to an odd address
$$$* THIS TRANSMISSION WAS MADE POSSIBLE BY 100% RECYCLED ELECTRONS
Received on Tue Nov 07 2000 - 11:37:35 NZDT