Interact hosting centers serve multiple service sites from a common hardware base. This paper presents the design and implementation of an architecture for resource management in a hosting center op-erating system, with an emphasis on energy as a driving resource management issue for large server clusters. The goals are to provi-sion server resources for co-hosted services in a way that automati-cally adapts to offered load, improve the energy efficiency of server dusters by dynamically resizing the active server set, and respond to power supply disruptions or thermal events by degrading service in accordance with negotiated Service Level Agreements (SLAs). Our system is based on an economic approach to managing shared server resources, in which services "bid " for resources as a func-tion of delivered performance. The system continuously moni-tors load and plans resource allotments by estimating the value of their effects on service performance. A greedy resource allocation algorithm adjusts resource prices to balance supply and demand, allocating resources to their most efficient use. A reconfigurable server switching infrastructure directs request traffic to the servers assigned to each service. Experimental results from a prototype confirm that the system adapts to offered load and resource avail-ability, and can reduce server energy usage by 29 % or more for a typical Web workload. 1.

Abstract not found

Bid-prices are becoming an increasingly popular method for controlling the sale of inventory in revenue management applications. In this form of control, threshold—or ‘‘bid’’—prices are set for the resources or units of inventory (seats on flight legs, hotel rooms on specific dates, etc.) and a product (a seat in a fare class on an itinerary or room for a sequence of dates) is sold only if the offered fare exceeds the sum of the threshold prices of all the resources needed to supply the product. This approach is appealing on intuitive and practical grounds, but the theory underlying it is not well developed. Moreover, the extent to which bid-price controls represent optimal or near optimal policies is not well understood. Using a general model of the demand process, we show that bid-price control is not optimal in general and analyze why bid-price schemes can fail to produce correct accept/deny decisions. However, we prove that when leg capacities and sales volumes are large, bid-price controls are asymptotically optimal, provided the right bid prices are used. We also provide analytical upper bounds on the optimal revenue. In addition, we analyze properties of the asymptotically optimal bid prices. For example, we show they are constant over time, even when demand is nonstationary, and that they may not be unique.

Abstract: Current capacity planning practices based on heavy over-provisioning of power infrastructure hurt (i) the operational costs of data centers as well as (ii) the computational work they can support. We explore a combination of statistical multiplexing techniques to improve the utilization of the power hierarchy within a data center. At the highest level of the power hierarchy, we employ controlled underprovisioning and over-booking of power needs of hosted workloads. At the lower levels, we introduce the novel notion of soft fuses to flexibly distribute provisioned power among hosted workloads based on their needs. Our techniques are built upon a measurement-driven profiling and prediction framework to characterize key statistical properties of the power needs of hosted workloads and their aggregates. We characterize the gains in terms of the amount of computational work (CPU cycles) per provisioned unit of power – Computation per Provisioned Watt (CPW). Our technique is able to double the CPW offered by a Power Distribution Unit (PDU) running the e-commerce benchmark TPC-W compared to conventional provisioning practices. Over-booking the PDU by 10 % based on tails of power profiles yields a further improvement of 20%. Reactive techniques implemented on our Xen VMM-based servers dynamically modulate CPU DVFS states to ensure power draw below the limits imposed by soft fuses. Finally, information captured in our profiles also provide ways of controlling application performance degradation despite overbooking. The 95 th percentile of TPC-W session response time only grew from 1.59 sec to 1.78 sec—a degradation of 12%.

Virtual nesting is a popular capacity control strategy in network revenue management. (See Smith et al. [36].) In virtual nesting, products (itinerary-fare-class combinations) are mapped ("indexed") into a relatively small number of "virtual classes" on each resource (flight leg) of the network. Nested protection levels are then used to control the availability of these virtual classes; specifically, a product request is accepted if and only if its corresponding virtual class is available on each resource required. (See Talluri and van Ryzin [38] for a detailed discussion of virtual nesting and protection level controls.) Bertsimas and de Boer [8] recently proposed an innovative simulation-based optimization method for computing protection levels in a virtual nesting control scheme. In contrast to traditional heuristic methods, their approach more accurately approximates the true network revenues generated by the virtual nesting controls. However, because it is based on a discrete model of capacity and demand, the method has both computational and theoretical limitations. In particular, it uses first-difference estimates, which are computationally complex to calculate exactly. These gradient estimates are then used in a steepest ascent type algorithm, which, for discrete problems, has no guarantee of convergence.

The airline revenue management problem can be decomposed into four distinct but related sub-problems that are usually treated separately: demand forecasting, overbooking, capacity allocation and pricing. Over the last decades, much interest has been devoted to the overbooking and capacity allocation issues and, today, most major airlines rely on computerized tools to deal with these two sub-problems. Pricing, however, has received less attention, which can be explained by the technical and theoretical di#culties inherent to the implementation of a practical Pricing Decision Support System. In this paper, we present a new modelling approach that allows for the joint solution of the capacity allocation and pricing sub-problems faced by a major North American airline. Using predefined booking limits, the resulting model can also applied be used in a "pure" pricing context. Our approach is based on the bilevel programming paradigm, a special case of hierarchical mathematical optimization. This modelling technique makes it possible to take into account matters such as customer segmentation, behavior with regard to fares and other product attributes, and the interactions induced by overlapping routes, which are typical of modern air transportation networks.

Customer choice behavior, such as \buy-up " and \buy-down", is an important phenomenon in a wide range of industries. Yet there are few models or methodologies available to exploit this phenomenon within yield management systems. We make some progress on ¯lling this void. Speci¯cally, we develop a model of yield management in which the buyers ' behavior is modeled explicitly using a multi-nomial logit model of demand. The control problem is to decide which subset of fare classes to o®er at each point in time. The set of open fare classes then a®ects the purchase probabilities for each class. We formulate a dynamic program to determine the optimal control policy and show that it reduces to a dynamic nested allocation policy. Thus, the optimal choice-based policy can easily be implemented in reservation systems that use nested allocation controls. We also develop an estimation procedure for our model based on the expectation-maximization (EM) method that jointly estimates arrival rates and choice model parameters when no-purchase outcomes are unobservable. Numerical results show that this combined optimization-estimation approach may signi¯cantly improve revenue performance relative to traditional leg-based models that do not account for choice behavior.

Mean-Range Based Distribution-Free Procedures to Minimize ”Overage ” and ”Underage ” Costs We introduce and discuss several mean-range based distribution-free decision procedures to minimize several “overage ” and “underage ” cost functions. For a general cost function, we identify the most favorable distribution and the least favorable distribution associated with the random variable and determine the upper and lower bounds for the cost function. For the quadratic cost function, we recommend the min-max distribution-free decision. For the linear cost function, we identify the range of potential optimal solutions (decisions) and recommend a hybrid distribution-free decision that has serval favorable properties. Several numerical examples are provided to demonstrate the robustness of the proposed distributionfree decisions.

Revenue management models traditionally assume that future demand is unknown, but can be represented by a stochastic process or a probability distribution. Demand is however often difficult to characterize, especially in new or nonstationary markets. In this paper, we develop robust formulations for the capacity allocation problem in revenue management, using the maximin and the minimax regret criteria, under general polyhedral uncertainty sets. Our approach encompasses the following open-loop controls: partitioned booking limits, nested fare classes by origin-destination pairs, Displacement-Adjusted Virtual Nesting, and fixed bid prices. We also characterize the optimal booking policy under interval uncertainty; while partitioned booking limits are optimal under the maximin criterion, some nesting is desirable under the minimax regret criterion. Our numerical analysis reveals that robust controls can outperform the classical heuristics for network revenue management, while achieving the best performance in the worst case. Our models are scalable to solve practical problems, because they combine efficient solution methods (small mixed-integer and linear optimization problems) with very modest data requirements. 1.

Can airline Yield Management strategies be used to generate additional revenue from spare capacity in telecom networks? Pundits believe “yes”, based on several analogies between the industries such as, for instance, perishable inventory and negligible marginal cost of usage. However, no one has yet described how, one of the chief difficulties being the vastly different nature of airlines products and telecom services. Motivated to show how Operations Research can play a role in structuring this area, we: (i) argue that telecom Yield Management should be based on ’innovative ’ services explicitly designed to use only spare capacity, (ii) propose, borrowing from airlines, a framework to simplify related decision modeling, and (iii) demonstrate both our argument and the framework by articulating several ’innovative ’ telecom services and modeling them to varying degrees of depth. This thesis focuses only on the decision-making that will be required within a large infrastructure for operating new ’Yield Management ’ services. For each service, several decision variables can be considered to maximize revenue from available capacity, e.g. pricing, capacity limits and admission control, among others. Incorporating all such decisions in a single model usually leads to complicated