UMD Queue List

List of available queues for UMD system

List of queues for VO, ROLE

VO belle ilc g4med ops ops
ROLE non admin prod non admin prod non admin prod non admin prod
gridbelle_short × ○ ○ × × × × × × × × ×
gridbelle_middle ○ ○ ○ × × × × × × × × ×
gridbelle_long × ○ ○ × × × × × × × × ×
gridbelle_heavy × × ○ × × × × × × × × ×
gridilc_short × × × ○ ○ ○ × × × × × ×
gridilc_middle × × × ○ ○ ○ × × × × × ×
gridilc_long × × × ○ ○ ○ × × × × × ×
gridilc_heavy × × × ○ ○ ○ × × × × × ×
gridg4med_short × × × × × × × ○ ○ × × ×
gridg4med_middle × × × × × × ○ ○ ○ × × ×
gridg4med_long × × × × × × × ○ ○ × × ×
gridg4med_heavy × × × × × × × × ○ × × ×
gridops × × × × × × × × × ○ ○ ○
gridshort × × × × × × × × × × × ×
gridmiddle × × × × × × × × × × × ×
gridlong × × × × × × × × × × × ×
VO calice cdfj dteam geant4 fkppl.kisti.re.kr kagra ppj vo.france-asia.org
ROLE non admin prod
gridbelle_short × × ×
gridbelle_middle × × ×
gridbelle_long × × ×
gridbelle_heavy × × ×
gridilc_short × × ×
gridilc_middle × × ×
gridilc_long × × ×
gridilc_heavy × × ×
gridg4med_short × × ×
gridg4med_middle × × ×
gridg4med_long × × ×
gridg4med_heavy × × ×
gridops × × ×
gridshort × ○ ○
gridmiddle ○ ○ ○
gridlong × ○ ○

non : no ROLE is assigned
admin : set Role=lcgadmin
prod : set Role=production

LSF parameters for GRID queue

Parameter Details gridbelle_short gridbelle_middle gridbelle_long gridbelle_heavy gridbelle_merge
PRIORITY Priority of queue 120 100 100 100 100
NICE value of nice 5 5 5 5 15
HJOB_LIMIT slot limit per host 10 20 26 10 40
UJOB_LIMIT slot limit per user 1 - - - - -
TASKLIMIT available slots per job 1 1 1 1 1
RES_REQ Resource requirements used to determine eligible hosts. affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core]
CPULIMIT cputime limit 3:00 12:00 24:00 168:00 168:00
RUNLIMIT walltime limit 6:00 24:00 48:00 240:00 240:00
MEMLIMIT memory limit per job [MB] 4096 4096 4096 4096 4096
SWAPLIMIT virtual memory limit per job [MB] - - - - -
PROCESSLIMIT processes per job 2 30 30 30 30 30
FAIRSHARE fairshare - - - - -
FAIRSHARE_QUEUES slave queue of fairshare
RERUNNABLE auto-rerun of jobs YES YES YES YES
PRE_EXEC pre-exec shell before job submission - - - -
USERS allowed users for job submission belleprd,
bellesgm		 bellegrid,
belleprd,
bellesgm		 belleprd,
bellesgm		 belleprd belleprd,
bellesgm
HOSTS target host calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1 calccomputeIO
SLOT_RESERVE slot reserve - - - - -
CORELIMIT core file size limit per process [KB] - - - - -
STACKLIMIT stack segment limit per process [KB] - - - - -
JOB_SATRTER enviroment before job submission ulimit -v 4808000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 4096;
Parameter Details gridilc_short gridilc_middle gridilc_long girdilc_heavy
PRIORITY Priority of queue 120 100 100 100
NICE value of nice 5 5 5 5
HJOB_LIMIT slot limit per host 10 20 26 10
UJOB_LIMIT slot limit per user 1 - - - -
TASKLIMIT available slots per task 1 1 1 1
RES_REQ Resource requirements used to determine eligible hosts. affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core]
CPULIMIT cputime limit 3:00 12:00 24:00 168:00
RUNLIMIT walltime limit 6:00 24:00 48:00 240:00
MEMLIMIT memory limit per job [MB] 4096 4096 4096 4096
SWAPLIMIT virtual memory limit per job [MB] - - - -
PROCESSLIMIT processes per job 2 30 30 30 30
FAIRSHARE fairshare - - - -
FAIRSHARE_QUEUES slave queue of fairshare
RERUNNABLE auto-rerun of jobs YES YES YES YES
PRE_EXEC pre-exec shell before job submission - - - -
USERS allowed users for job submission ilcgrid,
ilcprd,
ilcsgm		 ilcgrid,
ilcprd,
ilcsgm		 ilcgrid,
ilcprd,
ilcsgm		 ilcgrid,
ilcprd,
ilcsgm
HOSTS target host calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1
SLOT_RESERVE slot reserve - - - -
CORELIMIT core file size limit per process [KB] - - - -
STACKLIMIT stack segment limit per process [KB] - - - -
JOB_SATRTER enviroment before job submission ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;
Parameter Details gridg4med_short gridg4med_middle gridg4med_long girdg4med_heavy
PRIORITY Priority of queue 120 100 100 100
NICE value of nice 5 5 5 5
HJOB_LIMIT slot limit per host 10 20 26 10
UJOB_LIMIT slot limit per user 1 - - - -
TASKLIMIT available slots per task 1 1 1 1
RES_REQ Resource requirements used to determine eligible hosts. affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core]
CPULIMIT cputime limit 3:00 12:00 24:00 168:00
RUNLIMIT walltime limit 6:00 24:00 48:00 240:00
MEMLIMIT memory limit per job [MB] 4096 4096 4096 4096
SWAPLIMIT virtual memory limit per job [MB] - - - -
PROCESSLIMIT processes per job 2 30 30 30 30
FAIRSHARE fairshare USER_SHARES
[default, 1]		 USER_SHARES
[default, 1]		 USER_SHARES
[default, 1]		 USER_SHARES
[default, 1]
FAIRSHARE_QUEUES slave queue of fairshare
DISPATCH_ORDER dispatch order of queue - - - -
RERUNNABLE auto-rerun of jobs YES YES YES YES
PRE_EXEC pre-exec shell before job submission - - - -
USERS allowed users for job submission g4medprd,
g4medsgm		 g4medgrid,
g4medprd,
g4medsgm		 g4medprd,
g4medsgm		 g4medprd
HOSTS target host calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1
REQUEUE_EXIT_VALUES exit code of requeue - - - -
SLOT_RESERVE slot reserve - - - -
CORELIMIT core file size limit per process [KB] - - - -
STACKLIMIT stack segment limit per process [KB] - - - -
JOB_SATRTER enviroment before job submission ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;
Parameter Details gridops gridshort gridmiddle gridlong
PRIORITY Priority of queue 120 120 100 100
NICE value of nice 5 5 5 5
HJOB_LIMIT slot limit per host 10 10 20 26
UJOB_LIMIT slot limit per user 1 - - - -
TASKLIMIT available slots per job 1 1 1 1
RES_REQ Resource requirements used to determine eligible hosts. affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core] affinity[core(1):cpubind=core]
CPULIMIT cputime limit 10 3:00 12:00 24:00
RUNLIMIT walltime limit 30 6:00 24:00 48:00
MEMLIMIT memory limit per job [MB] 4096 4096 4096 4096
SWAPLIMIT virtual memory limit per job [MB] - - - -
PROCESSLIMIT processes per job 2 30 30 30 30
FAIRSHARE fairshare - - - -
FAIRSHARE_QUEUES slave queue of fairshare
DISPATCH_ORDER dispatch order of queue - - - -
RERUNNABLE auto-rerun of jobs YES YES YES YES
USERS allowed users for job submission opsgrid,
opsprd,
opssgm		 caliceprd,
calicesgm,
cdfjprd,
cdfjsgm,
dteamprd,
dteamsgm,
fasgm,
fkpplprd,
fkpplsgm,
geant4sgm,
kagraprd,
kagrasgm,
ppjprd,
ppjsgm		 calicegrid,
caliceprd,
calicesgm,
cdfjgrid,
cdfjprd,
cdfjsgm,
dteamgrid,
dteamprd,
dteamsgm,
fagrid,
fasgm,
fkpplgrid,
fkpplprd,
fkpplsgm,
geant4grid,
geant4sgm,
kagragrid,
kagraprd,
kagrasgm,
ppjgrid,
ppjprd,
ppjsgm		 calicesgm,
cdfjprd,
cdfjsgm,
dteamprd,
dteamsgm,
fasgm,
fkpplprd,
fkpplsgm,
geant4sgm,
kagraprd,
kagraprd,
kagrasgm,
ppjprd,
ppjsgm
HOSTS target host calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2		 calccomputeT1,
calccomputeT2
REQUEUE_EXIT_VALUES exit code of requeue - - - -
SLOT_RESERVE slot reserve - - - -
CORELIMIT core file size limit per process [KB] - - - -
STACKLIMIT stack segment limit per process [KB] - - - -
JOB_SATRTER enviroment before job submission ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;		 ulimit -v 16384000;
ulimit -u 1500;

Attention : These queue can not used for job submittion from work server by bsub command. Please use these queue for gLite job submission command.

  1. slot limit 28 in one host per user apart from UJOB_LIMIT limitation. ↩↩↩↩

  2. OS limit 500 processes/user apart from LSF limitation. ↩↩↩↩