Dear Robert, I was wondering if I could truncate an optimization process and get a "partial" solution ("the best" the solver has already gotten so far).
I am currently trying to solve a MILP with over 192.000 variables and 281.000 constrains (after presolve elimination), using CPLEX.
Previously, I solved a very similar model, in the same machine and with the same solver, getting an optimal solution in over 10 minutes.
I made a slight modification to one restriction of the model, changed the size of two sets, and tried to solve it again. The result is a model very similar in size to those I've run before. However, to this moment AMPL has been running for roughly 28 Hours, and has not given any solution yet.
Is it possible the model has gotten "stucked"? Can I get any solution or report of the progress after truncating the process?
Any suggestion is highly appreciated.