Exploring service relocation and differentiation to improve survivability on operation of resilient optical cloud networks

Abstract

Optical clouds are the combination of optical transport networks and cloud computing, which allows the integrated management of both infrastructures in one controller element. In this paradigm, cloud services can be provisioned in an anycast fashion, i.e., only the source node asking for a service and the amount of IT resources are specified, while it is up to the cloud control/management system to select the most suitable destination datacenter (DC) node. During the cloud service provisioning process resiliency is crucial in order to guarantee continuous network operations also in the presence of failures. On one hand, a survivability strategy needs to be able to meet the availability requirements of each specific cloud service, while on the other hand it must be efficient in using backup resources. Service relocation (i.e., the ability to live re-allocate one provisioned service to another DC) is one of the new features that can be used in this new paradigm, but needs to be applied carefully given its associated overhead may overload the network and DCs. Current works in the literature consider a single service model that lack representation of the heterogeneity expected for cloud services. In this context, some disrupted connection properties can be considered to improve the network survivability during the restoration process, e.g., priorities and service remaining holding time. This thesis proposes a restoration-based survivability strategy, which combines the benefits of both cloud service relocation and service differentiation concepts. The former is used to enhance the restorability performance (i.e., the percentage of successfully restored cloud services) offered by restoration, while the latter ensures that critical services are given the proper consideration while backup resources are assigned. An ILP and a heuristic are presented in order to solve the proposed survivability strategy. The proposed strategies are evaluated considering a variety of different simulation scenarios. Results show that performance achieved by the proposed ILP and heuristic are close to the ones achieved when using protection strategies, but with the inherent benefits in terms of efficient use of resources offered by restoration-based approaches.