[torqueusers] Torque queues jobs while there are free slots

Elvedin Trnjanin trnja001 at umn.edu
Wed Feb 18 10:20:17 MST 2009


I've had issues where there were two nodes each running a job that takes 
one slot. All subsequent job submissions are queued because all of the 
nodes have jobs running, although they could have plenty of free slots 
open. Is this expected behavior out of Torque with my current settings? 
I know there is an option to make the jobs start running across nodes 
that that have free slots even though they may have jobs on them, but 
one of our applications takes up 90% of the cluster slots and requires 
dedicated nodes, so we want to avoid that.

Below is the result of "qmgr -c 'p s'"

create queue batch
set queue batch queue_type = Execution
set queue batch max_running = 200
set queue batch resources_default.neednodes = batch
set queue batch resources_default.nodes = 1
set queue batch resources_default.walltime = 01:00:00
set queue batch max_user_run = 10
set queue batch enabled = True
set queue batch started = True
#
# Create and define queue computeone
#
create queue computeone
set queue computeone queue_type = Execution
set queue computeone acl_hosts = boar01.oit.umn.edu
set queue computeone acl_hosts += compute01
set queue computeone resources_default.nodes = 1
set queue computeone resources_default.walltime = 100:00:00
set queue computeone enabled = True
set queue computeone started = True
#
# Set server attributes.
#
set server scheduling = True
set server managers = maui at boar01.oit.umn.edu
set server managers += swadm at boar01.oit.umn.edu
set server operators += maui at boar01.oit.umn.edu
set server operators += swadm at boar01.oit.umn.edu
set server default_queue = batch
set server log_events = 511
set server mail_from = adm
set server scheduler_iteration = 600
set server node_check_rate = 150
set server tcp_timeout = 6
set server pbs_version = 2.2.1



More information about the torqueusers mailing list