How to sequentially (qsub?) run jobs in atk (cluster system)?

[faxer92]

Dear QW Experts,

How to sequentially run jobs from test1.py to test10 in atk (qsub in cluster system)?   
for example,

~/libexec/ mpiexec.hydra -machinefile MachineFile -n 24 ~/bin/atkpython test1.py 2>&1 | tee test1.log

~/libexec/ mpiexec.hydra -machinefile MachineFile -n 24 ~/bin/atkpython test2.py 2>&1 | tee test2.log

......

~/libexec/ mpiexec.hydra -machinefile MachineFile -n 24 ~/bin/atkpython test10.py 2>&1 | tee test10.log

thank you so much and appreciate of your concern.

[Ulrik G. Vej-Hansen]

That is really more a feature of the queuing system on your cluster than ATK, but most queuing systems have a feature for that. It is usually controlled by a keyword along the lines of "depend" or "dependency". I suggest you search the documentation for your cluster for those words.

[faxer92]

My system is CentOS, and PBS manager is Torque, is there an example for queue jobs ?  really appreciate your patience in answering  my questions

[Ulrik G. Vej-Hansen]

If you search for "torque depend pbs" on Google, both of the first 2 hits explain the concept.