Maximizing Risk-aware Last-Minute Inter-Datacenter Backup with Progressive Disasters

Network failure caused by disasters results in communication disruption and huge data loss in backbone Data Center (DC) networks. To minimize such losses, in this paper, we consider a risk-aware dynamic backup model that significantly reduces data-loss in a time-varying network hit by a progressive disaster. We propose a mixed-integer linear program (MILP) for risk-aware backup maximization over a dynamic network hit by a progressive disaster. Further, we perform extensive numerical simulations to show that our proposed MILP performs significantly well compared to existing algorithms which only address emergency backup in static networks hit by stationary disasters.