I'm encountering a problem while running a sensitivity analysis for an optimization problem in Julia using JuMP and the HiGHS solver. Initially, the model execution time varies between 5 to 20 minutes for different input parameter combinations. However, for some combinations, the model runs indefinitely, exceeding 24 hours without reaching a solution. The process appears to be stuck at a single iteration, with the execution time frozen at 108.8 seconds (as indicated in the attached screenshot).
I'm uncertain whether this issue is related to RAM limitations or another factor. Monitoring the Task Manager during execution shows that there is still available RAM, suggesting the problem might not be memory-related. To isolate the issue, I attempted running the model with only the problematic parameter sets, but the problem persists, ruling out the possibility of cumulative errors from serial execution.
Has anyone experienced a similar issue with the HiGHS solver or within a Julia-JuMP context? Any insights on potential causes or solutions would be greatly appreciated.
Below are the steps I've taken so far without success:
- Monitoring RAM usage to check for possible memory leaks or limitations.
- Isolating the problematic parameter sets to identify if the issue is specific to certain data.
- Reviewing the solver's documentation for known issues or limitations with certain types of optimization problems.
Any advice on further troubleshooting steps or solutions would be immensely helpful.