Originally posted by: SystemAdmin

For a large MIP problem, I was using default hard disk space for temporary file storage (after out of memory error). I did not save the results anytime and the run was aborted after 3 days due to power failure. Now 100 GB temporary files still saved on hard disk.

My question: is it possible to restore results using these temporary files on hard disk? or can these files be used to save on computation time in any ways?
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

No, sorry, the files cannot be used to restart CPLEX :-(
The data in the files is only valid for the process that created them.
One way to save information during a run is CPX_PARAM_INTSOLFILEPREFIX. If this parameter is set then CPLEX will write all integer feasible solutions it finds to disk. So you can at least recover the integer feasible solutions after a crash.
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

Thanks Daniel!

1. is it possible to set the parameter CPX_PARAM_INTSOLFILEPREFIX from IDE .ops or OPL ? If yes, plz tell me how. Or only in C/C++....APIs

2. So we can warm start the next run by finding the best from the last recovered integer feasible solutions?
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

> 1. is it possible to set the parameter CPX_PARAM_INTSOLFILEPREFIX from IDE .ops or OPL ? If yes, plz tell me how. Or only in C/C++....APIs
>
Sorry, I am not an expert with the OPL IDE, maybe you can ask this question on the OPL Forum. I took a quick look at the settings page in an OPL model but could not find the parameter there. What should definitely work is adding an execute block:

execute PARAMS {
  cplex.params[2143] = "filenameprefix";
}

> 2. So we can warm start the next run by finding the best from the last recovered integer feasible solutions?
>
Yes, that is what you could try. Whether that is effective highly depends on the characteristics of your problem.
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

Thanks Daniel !!
#CPLEXOptimizers
#DecisionOptimization