[torqueusers] Cannot enforce nodes resource limitations

Adrien Leygue adrien.leygue at gmail.com
Wed Jan 4 10:04:16 MST 2006

Dear Torque users,

I have compiled and installed Torque 2.0.0p2 on our cluster (Xeon, 2
processors per nodes), and I am now in the process of configuring the
server and the queues.

I wish to have at least two execution queues: one for parallel jobs
and another for sequential jobs (different maximum wall-time and
cpu-time will be imposed on these queues). There is another route
queue which is the default queue.

I thought this could be done by configuring the nodes resources of the
queues, but apparently I am missing something.

NB: I'd like to keep the system simple and I do not wish to use Maui
at this point.

My server configuration looks like this:

# Create and define queue route
create queue route
set queue route queue_type = Route
set queue route route_destinations = main
set queue route route_destinations += para
set queue route enabled = True
set queue route started = True

# Create and define queue main (one node per job)
create queue main
set queue main queue_type = Execution
set queue main from_route_only = True
set queue main resources_max.nodes = 1
set queue main resources_min.nodes = 1
set queue main enabled = True
set queue main started = True

create queue para
set queue para queue_type = Execution
set queue para from_route_only = True
set queue para resources_max.nodes = 4
set queue para resources_min.nodes = 2
set queue para enabled = True
set queue para started = True

set server scheduling = True
set server default_queue = route
set server log_events = 511
set server mail_from = adm
set server query_other_jobs = True
set server resources_default.neednodes = 1
set server resources_default.nodect = 1
set server resources_default.nodes = 1
set server scheduler_iteration = 600
set server node_check_rate = 150
set server tcp_timeout = 6
set server node_pack = True

The script I use for testing has the following node request:
#PBS -l nodes=2

When the job is running, I get the following qstat output:

Job ID             Username Queue    Jobname    SessID NDS TSK Memory
Time  S Time
------------------ -------- -------- ---------- ------ --- --- ------
----- - -----
1.glass         leygue   main     test_scrip    --      2  --    -- 
96:00 R   --

The job is running on 2 nodes, from the queue main which has a
max.nodes of 1, and apparently, the node_pack option has not been
taken into account!

I'd really appreciate any suggestion on this issue.

Thanks a lot.

Adrien Leygue.

More information about the torqueusers mailing list