Exception raised when WAIT_TIME is exceeded during Gurobi optimization solve

[matthewgartenhaus-8451]

Summary

When running with gurobi_direct, I get a failure when I specify the WORK_LIMIT and the WORK_LIMIT gets exceeded

Steps to reproduce the issue

# example.py
import pyomo.environ as pyo
from pyomo.opt import SolverFactory
import numpy as np

# Number of cities
num_cities = 100

# Distance matrix (symmetric)
np.random.seed(42)  # For reproducibility
distances = np.random.randint(1, 100, size=(num_cities, num_cities))
distances = (distances + distances.T) // 2  # Make the matrix symmetric
np.fill_diagonal(distances, 0)  # Set diagonal to 0

# Create a Concrete Model
model = pyo.ConcreteModel()

# Sets
model.CITIES = pyo.RangeSet(num_cities - 1)
model.EDGES = pyo.Set(initialize=[(i, j) for i in range(num_cities) for j in range(num_cities) if i != j])

# Variables
model.x = pyo.Var(model.EDGES, within=pyo.Binary)

# Objective function
def objective_rule(model):
    return sum(model.x[i, j] * distances[i, j] for i, j in model.EDGES)
model.obj = pyo.Objective(rule=objective_rule, sense=pyo.minimize)

# Constraints
def degree_rule(model, i):
    return sum(model.x[i, j] for j in range(num_cities) if i != j) == 1
model.degree = pyo.Constraint(model.CITIES, rule=degree_rule)

def degree_rule_2(model, j):
    return sum(model.x[i, j] for i in range(num_cities) if i != j) == 1
model.degree_2 = pyo.Constraint(model.CITIES, rule=degree_rule_2)

# Subtour elimination constraints
model.u = pyo.Var(model.CITIES, within=pyo.NonNegativeIntegers, bounds=(0, num_cities-1))

def subtour_elimination_rule(model, i, j):
    if i != j and (i != 0 and j != 0):
        return model.u[i] - model.u[j] + model.x[i, j] * num_cities <= num_cities - 1
    return pyo.Constraint.Skip
model.subtour_elimination = pyo.Constraint(model.EDGES, rule=subtour_elimination_rule)

# Solve the model
solver = SolverFactory('gurobi_direct')
solver.options["WorkLimit"] = 1.0
results = solver.solve(model, tee=True)

Error Message

---

Explored 1 nodes (1344 simplex iterations) in 1.01 seconds (1.00 work units)
Thread count was 16 (of 16 available processors)

Solution count 10: 974 1008 1032 ... 1833

Work limit reached
Best objective 9.740000000000e+02, best bound 9.550000000000e+02, gap 1.9507%


---> 50 results = solver.solve(model, tee=True)

File .venv/lib/python3.10/site-packages/pyomo/solvers/plugins/solvers/direct_solver.py:198, in DirectSolver.solve(self, *args, **kwds)
    196 else:
    197     if self._load_solutions:
--> 198         _model.solutions.load_from(
    199             result,
    200             select=self._select_index,
    201             default_variable_value=self._default_variable_value,
    202         )
    203         result._smap_id = None
    204         result.solution.clear()

File .venv/lib/python3.10/site-packages/pyomo/core/base/PyomoModel.py:231, in ModelSolutions.load_from(self, results, allow_consistent_values_for_fixed_vars, comparison_tolerance_for_fixed_vars, ignore_invalid_labels, id, delete_symbol_map, clear, default_variable_value, select, ignore_fixed_vars)
    226         logger.warning(
    227             "Loading a SolverResults object with "
    228             "an 'aborted' status, but containing a solution"
    229         )
    230     else:
--> 231         raise ValueError(
    232             "Cannot load a SolverResults object "
    233             "with bad status: %s" % str(results.solver.status)
    234         )
    235 if clear:
    236     #
    237     # Clear the solutions, but not the symbol map
    238     #
    239     self.clear(clear_symbol_maps=False)

ValueError: Cannot load a SolverResults object with bad status: error

Information on your system

Pyomo version: 6.8.2
Python version: 3.10.16
Operating system: 15.1.1
How Pyomo was installed (PyPI, conda, source): PyPI
Solver (if applicable): Gurobi Compute Server

Additional information

It is unclear to me why this is failing based on this code:

pyomo/pyomo/solvers/plugins/solvers/GUROBI_RUN.py

Lines 176 to 180 in 038f38b

	elif hasattr(GRB, "WORK_LIMIT") and (solver_status == GRB.WORK_LIMIT):
	status = 'aborted'
	message = 'Optimization terminated because the work expended exceeded the value specified in the WorkLimit parameter.'
	term_cond = 'maxTimeLimit'
	solution_status = 'stoppedByLimit'