Losing Residual Datas
 

Hello everbody,

I' m running an analysis with StarCD in batch mode (using Linux). When I want to change RFs, I` m terminating the analysis (with kill command). But, in this conditon after the restart I' m always losing my residual datas. Does anybody know a solution for this problem? Or may be a command?

Thanks, Tunc

Re: Losing Residual Datas
 

If you terminate a parallel run with star -kill, your *.rsi file is not written out properly (i.e. into your working directory). You could either stop your run with star -abort (the recommended way to stop) or you have a look in the first parallel sub-directory. You should find a *.rsi file in there.

Re: Losing Residual Datas
 

You can change RFs without aborting the run. Set switch 50 and modify the corresponding .ctrl file(s). With switch 50 set on .ctrl is read after every iteration/timestep

Regards, Oliver

Re: Losing Residual Datas
 

Thanks Oliver. But under which location can I find the .ctrl file(s)?

Regards, Tunc

Re: Losing Residual Datas
 

The .ctrl files are usually in the working directory. They are created during startup of calculation. Which version and which platform do you use?

Regards, Oliver

Re: Losing Residual Datas
 

I' m using version 3.24 and my platform is Suse Linux 9.

Regards, Tunc

Re: Losing Residual Datas
 

I use the same platform but already Star 3.26. The ctrl file is - as I already wrote - located in your calculation directory (where .geom, .prob, etc can be found) or for parallel calculations in directory _0001.

It is more or less the same mechanism as used from starwatch. Up to 3.15 it was also documented in User Guide, but it is still working with 3.2x

Regards, Oliver

Re: Losing Residual Datas
 

Thank you Oliver.

Regards, Tunc