TORQUE资源管理器不按正常顺序排队作业

时间:2017-08-08 18:31:26

标签: queue torque

我有扭矩的奇怪行为。我设置了一个独特的队列来扭转每个作业应该具有相同的优先级,但是当我提交其中一些时,它们似乎是随机运行的。 要提交它们,我使用此命令qsub -q prod script.sh

例如,使用qstat命令我有这个。

4566.localhost             ..._1.PBS_1_2.sh thomas.pedot           0 Q prod
4567.localhost             ..._1.PBS_1_3.sh thomas.pedot           0 Q prod
4568.localhost             ..._1.PBS_1_4.sh thomas.pedot    20:16:34 C prod
4569.localhost             ..._1.PBS_1_5.sh thomas.pedot    17:55:30 R prod
4570.localhost             ..._1.PBS_1_6.sh thomas.pedot           0 Q prod

为什么职位编号4568和4569在4566 4567等之前运行?

我想念什么?

更新: 以前的工作中qstat -f的结果:

Job_Name = SCRIPT.sh
Job_Owner = thomas.pedot@localhost
job_state = Q
queue = prod
server = localhost
Checkpoint = u
ctime = Wed Aug  9 11:09:40 2017
Error_Path = xxxx:/dev/null
Hold_Types = n
Join_Path = n
Keep_Files = n
Mail_Points = a
mtime = Wed Aug  9 11:09:40 2017
Output_Path = xxxx:/dev/null
Priority = 0
qtime = Wed Aug  9 11:09:40 2017
Rerunable = True
Resource_List.neednodes = 1
Resource_List.nodect = 1
Resource_List.nodes = 1
Resource_List.walltime = 168:00:00
substate = 10
Variable_List = xxxx
euser = thomas.pedot
egroup = users
queue_rank = 606
queue_type = E
comment = Not Running: Queue job limit has been reached.
etime = Wed Aug  9 11:09:40 2017
submit_args = -q prod -e /dev/null -o /dev/null SCRIPT.sh
fault_tolerant = False
job_radix = 0
submit_host = localhost

我认为以后的工作没有什么不同。

非常感谢, 托马斯

0 个答案:

没有答案