Data-driven mixed-Integer linear programming-based optimisation for efficient failure detection in large-scale distributed systems
Abstract: Highlights•A nonlinear optimisation model for high-performing probabilistic failure detectors.•Transformed nonlinear optimisation model to Mixed-Integer Linear Problem.•New self-optimised failure detector using network data and the model optimal solutions.•Design of a heuristic algorithm to efficiently tackle the scalability challenge.•Testbeds development on both the Amazon Cloud and a Cloud simulator for robust results.
External IDs:dblp:journals/eor/RahmadiM22
Loading