it is not efficient to just wait for the job to finish.
utilize more nodes to speedup the development.
1 . use supermicro24
just copy a directory and ure fine.
2 . use Lucy cluster
here i need to rewrite the code to torque not gnuparallel. it won't be so much of effort though.
what is most important is to find where is the problem. not sure how to find errors effectively.
need a way to dump why the execution didn't work as intended.
what about outputting keywords to signal where the execution is on.
this way we can see where it failed.
it is not efficient to just wait for the job to finish. utilize more nodes to speedup the development.
1 . use supermicro24 just copy a directory and ure fine.
2 . use Lucy cluster here i need to rewrite the code to torque not gnuparallel. it won't be so much of effort though.
what is most important is to find where is the problem. not sure how to find errors effectively. need a way to dump why the execution didn't work as intended. what about outputting keywords to signal where the execution is on. this way we can see where it failed.