[GE users] Queue priority strangeness.

adary adary at marvell.com
Tue Aug 18 10:56:51 BST 2009


Is there a difference between the long job and a regular job?

Basically, its very easy to write a script that will find the process ID of your job and renice it to 0 (cant go lower if you are not root)

We had issues with our users being clever and doing a similar thing and the only solution was to yell^H^H^H^Htalk to them and ask nicely not to do something similar.



-----Original Message-----
From: udowaechter [mailto:udo.waechter at uni-osnabrueck.de]
Sent: Tuesday, August 18, 2009 12:51 PM
To: users at gridengine.sunsource.net
Subject: [GE users] Queue priority strangeness.

Hello,
I have a strange problem with one of our queues on GE 6.2u3.

We have recently defined a queue containing a subset of our machines
that are in the main queue. This should be the longrun queue containing
those machines that are guaranteed to run for long times.
Anyway, the problem is, that the jobs in this queue all run wiht "nice
0" although  it should have "nice 10"

All other queues' priority is honored.

How could I further debug this problem? Did anyone else experience this
problem?


Thanks,
udo.

Here is the config of the two queues:

1st, working priorities.


qname                 ikw
hostlist              @allhosts_ikw-slots_1 @allhosts_ikw-slots_2 \
                       @allhosts_ikw-slots_4 @allhosts_ikw-slots_8
seq_no                0
load_thresholds       np_load_avg=1.75
suspend_thresholds    NONE
nsuspend              1
suspend_interval      00:05:00
priority              7
min_cpu_interval      00:05:00
processors            UNDEFINED
qtype                 BATCH
ckpt_list             NONE
pe_list               make
rerun                 FALSE
slots
1,[@allhosts_ikw-slots_1=1],[@allhosts_ikw-slots_2=2], \
                       [@allhosts_ikw-slots_4=4],[@allhosts_ikw-slots_8=8]
tmpdir                /work/tmp
shell                 /bin/bash
prolog                NONE
epilog                NONE
shell_start_mode      posix_compliant
starter_method        NONE
suspend_method        NONE
resume_method         NONE
terminate_method      NONE
notify                00:00:60
owner_list            NONE
user_lists            GE-users
xuser_lists           ikw nkg www-data
subordinate_list      NONE
complex_values        NONE
projects              NONE
xprojects             NONE
calendar              NONE
initial_state         enabled
s_rt                  INFINITY
h_rt                  INFINITY
s_cpu                 INFINITY
h_cpu                 INFINITY
s_fsize               INFINITY
h_fsize               INFINITY
s_data                INFINITY
h_data                INFINITY
s_stack               INFINITY
h_stack               INFINITY
s_core                INFINITY
h_core                INFINITY
s_rss                 INFINITY
h_rss                 INFINITY
s_vmem                INFINITY
h_vmem                INFINITY


2nd queue, nice value not working:


qname                 ikw_longrun
hostlist              @allhosts_ikw_longrun-slots_2 \
                       @allhosts_ikw_longrun-slots_4 \
                       @allhosts_ikw_longrun-slots_8
seq_no                0
load_thresholds       np_load_avg=1.75
suspend_thresholds    NONE
nsuspend              1
suspend_interval      00:05:00
priority              10
min_cpu_interval      00:05:00
processors            UNDEFINED
qtype                 BATCH
ckpt_list             NONE
pe_list               make
rerun                 FALSE
slots                 1,[@allhosts_ikw_longrun-slots_2=2], \
                       [@allhosts_ikw_longrun-slots_4=4], \
                       [@allhosts_ikw_longrun-slots_8=8]
tmpdir                /work/tmp
shell                 /bin/bash
prolog                NONE
epilog                NONE
shell_start_mode      posix_compliant
starter_method        NONE
suspend_method        NONE
resume_method         NONE
terminate_method      NONE
notify                00:00:60
owner_list            NONE
user_lists            GE-users
xuser_lists           ikw nkg www-data
subordinate_list      NONE
complex_values        NONE
projects              NONE
xprojects             NONE
calendar              NONE
initial_state         enabled
s_rt                  INFINITY
h_rt                  INFINITY
s_cpu                 INFINITY
h_cpu                 INFINITY
s_fsize               INFINITY
h_fsize               INFINITY
s_data                INFINITY
h_data                INFINITY
s_stack               INFINITY
h_stack               INFINITY
s_core                INFINITY
h_core                INFINITY
s_rss                 INFINITY
h_rss                 INFINITY
s_vmem                INFINITY
h_vmem                INFINITY

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=212810

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=212811

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list