Difference between revisions of "QMUL Multicore Gridengine Configuration"

From GridPP Wiki
Jump to: navigation, search
(Created page with " Create parallel environment (PE) with dynamic cores PE configuration and attach this to th equeue that will run the multicore jobs qconf -sp smp, (man sge_pe for mor info...")
 
 
(4 intermediate revisions by one user not shown)
Line 1: Line 1:
 +
Create parallel environment (PE) with dynamic cores PE configuration and attach this to the queue that will run the multicore jobs
  
 +
For more info see https://twiki.cern.ch/twiki/bin/view/LCG/SGE
  
 +
qconf -sp smp, (man sge_pe for more info)
  
 +
pe_name            smp
 +
slots              2448   
 +
user_lists        NONE
 +
xuser_lists        NONE
 +
start_proc_args    /bin/true
 +
stop_proc_args    /bin/true
 +
allocation_rule    $pe_slots
 +
control_slaves    FALSE
 +
job_is_first_task  TRUE
 +
urgency_slots      min
 +
accounting_summary FALSE
  
Create parallel environment (PE) with dynamic cores PE configuration and attach this to th equeue that will run the multicore jobs
+
Reservations (draining nodes for multi core jobs) needs to be enabled. '-R y' flag enabling reservations is not added by 'sge_submit.sh' current workaround is running 'qalter -R y $list_of_mc_jobs' by a cron.
  
qconf -sp smp, (man sge_pe for mor info)
+
#!/bin/bash
 +
# reserve slots for jobs requesting 8 cores
 +
. /etc/profile.d/sge.sh
 +
/usr/bin/qstat -u \* | grep " 8 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
 +
/usr/bin/qstat -u \* | grep " 2 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
 +
/usr/bin/qstat -u \* | grep " 4 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
  
pe_name            smp
 
slots              2448   
 
user_lists        NONE
 
xuser_lists        NONE
 
start_proc_args    /bin/true
 
stop_proc_args    /bin/true
 
allocation_rule    $pe_slots
 
control_slaves    FALSE
 
job_is_first_task  TRUE
 
urgency_slots      min
 
accounting_summary FALSE
 
  
    reservations (draining nodes for multi core jobs) need to be enabled in addition
+
While you are at it why not enable mpi, you then also get the mpi nagios check for free. note you need to make sure that the PE SMP is listed first in order to make sure the multicore job is run only on one node.
        '-R y' flag enabling reservations is not added by 'sge_submit.sh'
+
        current workaround is running 'qalter -R y $list_of_mc_jobs' by a cron
+
  
#!/bin/bash
+
https://wiki.egi.eu/wiki/MAN03_MPI-Start_Installation_and_Configuration
# reserve slots for jobs requesting 8 cores
+
. /etc/profile.d/sge.sh
+
/usr/bin/qstat -u \* | grep " 8 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
+
/usr/bin/qstat -u \* | grep " 2 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
+
/usr/bin/qstat -u \* | grep " 4 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
+
  
 
+
[[Category: Multicore]]
 
+
https://twiki.cern.ch/twiki/bin/view/LCG/SGE
+

Latest revision as of 11:25, 11 February 2016

Create parallel environment (PE) with dynamic cores PE configuration and attach this to the queue that will run the multicore jobs

For more info see https://twiki.cern.ch/twiki/bin/view/LCG/SGE

qconf -sp smp, (man sge_pe for more info)

pe_name            smp 
slots              2448    
user_lists         NONE
xuser_lists        NONE
start_proc_args    /bin/true
stop_proc_args     /bin/true
allocation_rule    $pe_slots
control_slaves     FALSE
job_is_first_task  TRUE
urgency_slots      min
accounting_summary FALSE

Reservations (draining nodes for multi core jobs) needs to be enabled. '-R y' flag enabling reservations is not added by 'sge_submit.sh' current workaround is running 'qalter -R y $list_of_mc_jobs' by a cron.

#!/bin/bash
# reserve slots for jobs requesting 8 cores
. /etc/profile.d/sge.sh
/usr/bin/qstat -u \* | grep " 8 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
/usr/bin/qstat -u \* | grep " 2 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'
/usr/bin/qstat -u \* | grep " 4 " | awk '{ system("/usr/bin/qalter " $1 " -R y")}'


While you are at it why not enable mpi, you then also get the mpi nagios check for free. note you need to make sure that the PE SMP is listed first in order to make sure the multicore job is run only on one node.

https://wiki.egi.eu/wiki/MAN03_MPI-Start_Installation_and_Configuration