Originally posted by: SystemAdmin

I am trying to understand what the following error implies, and how to resolve the issue. I am solving a mixed-integer linear program with around 20,000 variables and 50,000 constraints and 100,000 non-zeros using Concert Technology and Visual C++. CPLEX was running for some 4 or 5 hours before an error occurred, and I want to understand why it happened so that I can fix it.

Does this have to do with memory used to store the tree? If so, which is more important - the actual tree size, or the compressed size? Is there any significance to the last node number being 250,000?

In any case, it would be great to have some more insight into possible ways to get around this issue.

Here are the last few rows of the log:

249900 224216 2.6562 98 3.8840 0.0000 3213052 100.00%

250000 224298 0.0000 190 3.8840 0.0000 3214084 100.00%

Elapsed time = 17783.28 sec. (tree size = 1788.90 MB, solutions = 17)
Nodefile size = 1661.55 MB (307.63 MB after compression)
Compressing row and column files.
Basis recovery file written to 'nodelp.xx'.

Implied bound cuts applied: 82
Flow cuts applied: 94
Mixed integer rounding cuts applied: 34
Gomory fractional cuts applied: 79

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Press any key to continue . . .
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

Yes, the memory from the tree is the problem. You can send the nodefile to disk instead of keeping it in memory by setting the NodeFileInd parameter. Look here for some tips: Running out of memory
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

Oops, bad link, should have tested before posting. Try this: Running out of memory
#CPLEXOptimizers
#DecisionOptimization

Originally posted by: SystemAdmin

Thanks for the link Mary, it's good to know that my C++ code wasn't the likely culprit, but memory limitations. I'm going to start with the strong branching option, and see how that goes.
#CPLEXOptimizers
#DecisionOptimization