`SolverResults Error` When Parallelising Pyomo Optimisations

I’m trying to optimise multiple linear programming problems in parallel using Pyomo and the standard Python multiprocessing library. When switching to using multi-processing I keep running into the error: ValueError: Cannot load a SolverResults object with bad status: error.

A similar issue was reported in this question, where their problem seemed to be that the solver (n.b. they used cbc whereas I used cplex) was timing out and couldn’t gracefully quit. This doesn’t seem to be the issue with my error though. One guess is that the cplex solver is trying to write to a temporary file that is then getting overwritten by the parallel optimisations.

Any help fixing this error would be much appreciated! The following code should reproduce the error.

import time  
import multiprocessing  
import pyomo.environ as pyo  
from pyomo.opt import SolverFactory  
  
def run_model(n_runs=50):  
    for _ in range(n_runs):  
        opt = SolverFactory(solver_name, executable=solver_executable)  
  
        model = pyo.ConcreteModel()  
        model.x = pyo.Var([1,2], domain=pyo.NonNegativeReals)  
        model.OBJ = pyo.Objective(expr = 2*model.x[1] + 3*model.x[2])  
        model.Constraint1 = pyo.Constraint(expr = 3*model.x[1] + 4*model.x[2] >= 1)  
  
        opt.solve(model)  
        time.sleep(0.1)  
  
    return  
  
  
solver_name = 'cplex'  
solver_executable = 'C:/Program Files/IBM/ILOG/CPLEX_Studio201/cplex/bin/x64_win64/cplex'  
  
  
if __name__ == '__main__':  
  
    process_list = []  
  
    for _ in range(10):  
        p = multiprocessing.Process(target=run_model)  
        p.start()  
        process_list.append(p)  
  
    for process in process_list:  
        process.join()

Process Process-8:
Traceback (most recent call last):
  File "C:UsersUseranaconda3envsenvlibmultiprocessingprocess.py", line 315, in _bootstrap
    self.run()
  File "C:UsersUseranaconda3envsenvlibmultiprocessingprocess.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "C:UsersUserDesktoplibrary_dirscriptsmultiprocessing_test.py", line 15, in run_model
    opt.solve(model)
  File "C:UsersUseranaconda3envsenvlibsite-packagespyomooptbasesolvers.py", line 626, in solve
    _model.solutions.load_from(
  File "C:UsersUseranaconda3envsenvlibsite-packagespyomocorebasePyomoModel.py", line 224, in load_from
    raise ValueError("Cannot load a SolverResults object "
ValueError: Cannot load a SolverResults object with bad status: error

Answer

A solution to the overall problem (but not this specific error) was to use glpk solver from here.

Advertisement

Answer