Time-varying resilient virtual network mapping for multi-location cloud data centers

Abstract

Optical networks constitute a fundamental building block that has enabled the success of cloud computing. Virtualization, a cornerstone of cloud computing, today is applied in the networking field: physical network infrastructure is logically partitioned into separate virtual networks, thus providing isolation between distinct virtual network operators (VNOs). Hence, the problem of virtual network mapping has arisen: how to decide which physical resources to allocate for a particular virtual network? In a cloud context, not just network connectivity is required, but also data center (DC) resources located at multiple locations, for computation and/or storage. Given the underlying anycast routing principle, the network operator has some freedom to which specific DC to allocate these resources.In this paper, we solve a resilient virtual network mapping problem that optimally decides on the mapping of both network and multi-location data center resources resiliently using anycast routing, considering time-varying traffic conditions. In terms of resilience, we consider the so-called VNO-resilience scheme, where resilience is provided in the virtual network layer. To minimize physical resource capacity requirements, we allow reuse of both network and DC resources. The failures we protect against include both network and DC resource failures: we hence allocate backup DC resources, and also account for synchronization between primary and backup DC.As optimization criteria, we not only consider resource usage minimization, but also aim to limit virtual network reconfigurations from one time period to the next. We propose a scalable column generation approach to solve the dynamic resilient virtual network mapping problem, and demonstrate it in a case study on a nationwide US backbone network.

A scalable model for multi-period virtual network mapping for resilient multi-site data centers

“…The VNO-resilience model we adopt is exactly the same as in our earlier work [6], [8] and illustrated in Fig. 1 Further, we assume that there is an automatic switchback to the original network path and DC once a fault is repaired, and therefore we allow reusing the same network/DC capacity to protect against other failures: backup capacity is shared.…”

“…As in [6], we investigate whether it is worth reconfiguring the primary and the backup paths in order to save bandwidth when the communication traffic pattern changes. Note that this change is not necessarily limited to a scaling of the volume, but also its geographical pattern/distribution: large backbone networks (such as the ones that we are designing VNets over) might comprise different time zones where activities are shifted in time, and hence the resulting volume of cloud requests fluctuates differently.…”

“…But these works considered "classical" traffic, i.e., not the cloud computing scenario with anycast routing. The current paper builds on our earlier work [6], which to our knowledge was the first to consider resilient multi-period anycast traffic routing. Yet, there we assumed an iterative approach: we formulated the optimization problem to find the best routing going from one period to the next.…”

Exploring the Benefit of Rerouting Multi-Period Traffic to Multi-Site Data Centers [Invited]

“…As an example, the first case in (11) enforces that the amount of traffic from source node v carried over path π does not decrease from period t − 1 to period t if the total traffic increases (i.e., if ∆ v,t ≥ ∆ v,t−1 ): if we have a volume x on that path for source v in period t − 1, at least the same volume will still cross it during t.…”

“…To the best of our knowledge, [11] was the first to study resilient multi-period anycast traffic routing. 1 Still, that work adopted an iterative approach, solving a single transition from one period to the next.…”