CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   STAR-CCM+ (http://www.cfd-online.com/Forums/star-ccm/)
-   -   Benchmark Commannd Line (http://www.cfd-online.com/Forums/star-ccm/112369-benchmark-commannd-line.html)

eRzBeNgEl January 27, 2013 09:58

Benchmark Commannd Line
 
Hi guys,

I just want to benchmark our new hpc cluster and need information about the command line how to start starccm by console.

I used in earlier versions the following command:
starccm+ benchmark.sim -benchmark:" -nits 1000 -nps 1,2,3,4,5,6" -np 6

Somehow it does not work anymore. starccm starts the process always with kust one node and it says that the command extenstion "-benchmark" does not exist. But it worked with Star-CCM+ V6xxx

My current version is V7.06.009 and my OS is Debian Linux.

Thanks for help

murash-v February 13, 2013 05:22

Hi. I have the same problem now. Did you solve it?

Quote:

Originally Posted by eRzBeNgEl (Post 404324)
Hi guys,

I just want to benchmark our new hpc cluster and need information about the command line how to start starccm by console.

I used in earlier versions the following command:
starccm+ benchmark.sim -benchmark:" -nits 1000 -nps 1,2,3,4,5,6" -np 6

Somehow it does not work anymore. starccm starts the process always with kust one node and it says that the command extenstion "-benchmark" does not exist. But it worked with Star-CCM+ V6xxx

My current version is V7.06.009 and my OS is Debian Linux.

Thanks for help


eRzBeNgEl February 17, 2013 16:27

Hi
You need just to wait. The benchmark starts with one node, iterates e.g. 100xtimes and then it will benchmark with two nodes up to max level you have entered in "-nps". I think there is a command extension where you can say with which node the benchmark should start. The "-benchmark" option itself only worked on one machine. If i wanted to check two machines e.g. "-on cluster01:6, cluster02:6" the benchmark mode did not work. Therefore I just created a few reports (Total overall elapsed time, cpu time) and run a "normal" simulation. For results I piped the ouput window into a log file and made my checks.


All times are GMT -4. The time now is 00:23.