Optimal network locality in distributed virtualized data-centers

Abstract

Cost efficiency is a key aspect in deploying distributed service in networks within decentralized service delivery architectures. In this paper, we address this aspect from an optimization and algorithmic standpoint. The research deals with the placement of service components to network sites, where the performance metric is the cost for acquiring components between the sites. The resulting optimization problem, which we refer to as the k-Component Multisite Placement Problem, is applicable to service distribution in a wide range of communication networking scenarios. We provide a theoretical analysis of the problem's computational complexity, and develop an integer programming model for providing reference results for performance benchmarking. On the algorithmic side, we present four approaches: an algorithm with approximation guarantee and three heuristics algorithms. The first heuristic is derived from graph theory on domatic partition. The second heuristic, built on intuition, admits distributed computation. The third heuristic emphasizes on fairness in cost distribution among the sites. We report simulation results for sets of networks where cost is represented by round-trip time (RTT) originating from real measurements. For small networks, the integer model is used to study algorithm performance in terms of optimality. Large networks are used to compare the al- * Corresponding author1 The work of this author is supported by CENIIT, Linköping University, Sweden 2 The work of this author is supported by Vipeer project and Agence Nationale de la Recherche, France February 11, 2011 gorithms relatively to each other. Among the algorithms, the heuristic based on intuition has close-to-optimal performance, and the fairness heuristic achieves a good balance between single-site cost and the overall one. In addition, the experiments demonstrate the significance of optimization for cost reduction in comparison to a the random allocation strategy. Preprint submitted to Elsevier

Cost-Minimizing Dynamic Migration of Content Distribution Services into Hybrid Clouds

“…Theorem 3: (Optimality of Cost) Let M k denote the optimal objective function value in the T-slot Lookahead problem (22) in time frame F k . The minimum operational cost derived with our algorithm is M (t) in time slot t. Suppose the time lasts for KT time slots, where K is a constant.…”

“…Related service placement problems: Placement of services onto different sites has been investigated [21] [22] based on the theories of Facility Location Problems (FLP) [23], including the k-Median Problem (kMP) [24] and k-Component Multi-Site Placement Problem (k-CMSP) [22]. Such a problem typically involves an NP-hard integer program, and can only be solved by approximation algorithms; it focuses on one-time optimization with fixed service demands, rather than online optimization over a long run of the system.…”

“…In order to solve (22) which contains non-linear items [y (22) to an equivalent problem as follows: To derive the optimal solution, we use an open source tool MOSEK [35]. We find that it is very time consuming to solve the T-slot lookahead problem in (32) using the MOSEK solver in our default system scale, even when we are merely optimizing the decisions in T = 6 time slots.…”

Routing-policy aware peering for large content providers

“…For example, in [9], a constant approximation distributed algorithm was proposed to solve the general data placement problem. Several works have also investigated distributed optimization of the service placement problem [40,41,47] in order to solve the scalability issue caused by global knowledge requirements. These studies differ from ours in that our model takes routing policies into consideration when selecting content peers.…”

“…On the surface, our problem might be solved with content placement algorithms [9,21,34,40,41,47,51,58]. However, the fact that routing policies and peering agreements are taken into account in the selection of peers makes the problem much more challenging.…”