A Multi-Optimizer Metaheuristic Approach for Solving Constrained Patch Scheduling Problem

Timely security patches are crucial in reducing vulnerabilities across computing infrastructures, yet excessive downtime or violations of concurrency constraints can adversely affect service availability and overall performance. In this work, we optimize the problem of constrained patch scheduling using a multi-optimizer metaheuristic approach. We first model patch scheduling as an optimization problem aiming to minimize both total downtime and vulnerability exposure while also respecting concurrency limits, which restrict the number of machines simultaneously offline. The objective function incorporates a penalty factor to discourage over-concurrent patching. We compare eight modern metaheuristic algorithms with variable patch durations. Extensive experiments are conducted, analyzing best objective scores, downtime, exposure, concurrency penalty, computational time, and convergence trends. **Our findings indicate that Equilibrium Optimizer (EO) achieves the lowest final cost (6.5351), followed closely by Multi-Verse Optimizer (MVO) (6.8295). Whereas Grey Wolf Optimizer (GWO) and Moth Flame Optimization (MFO), yield moderate results, while Whale Optimization Algorithm (WOA) exhibits the highest cost (10.3944). These outcomes underscore the impact of concurrency-limited patching constraints on performance and emphasize the benefits of selecting optimizers that effectively balance exploration and exploitation for scheduling in security-critical environments.