Delay-Aware Backup Resource Allocation with Probabilistic Protection for Network Services

This paper proposes a backup resource allocation model for virtual network functions (VNFs) to minimize the total required backup computing capacity with considering the service delay. If random failures occur to primary hosts, the VNFs in failed hosts are recovered by backup hosts, where the allocation is determined in advance. We introduce the probabilistic protection, where the probability that the protection provided by a backup host fails is limited within a given value; it allows backup resource sharing to reduce the total required computing capacity. The previous work formulated the backup resource allocation problem without considering the service delay as a mixed integer linear programming (MILP) problem by adopting the robust optimization. We consider the delay of services, which consists of networking delay between hosts and processing delay in each requested VNF. The probability that the total delay of a service exceeds its threshold is constrained within a given value. To solve the problem with the delay constraint, we introduce an algorithm with two methods to make the MILP problem be aware of the service delay. The results observe that, compared to the baseline, the proposed model can reduce the total required backup capacity of computing resource.