scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

On Economic and Computational-Efficient Resource Pricing in Large Distributed Systems

17 May 2010-pp 838-843
TL;DR: This paper proposes a dynamic pricing mechanism for the allocation of shared resources, and performs both theoretical and simulation analysis to evaluate the economic and computational efficiency of the allocation and the scalability of the mechanism.
Abstract: There is growing interest in large-scale systems where globally distributed and commoditized resources can be shared and traded, such as peer-to-peer networks, grids, and cloud computing. Users of these systems are rational and maximize their own interest when consuming and contributing shared resources, even if by doing so they affect the overall efficiency of the system. To manage rational users, resource pricing and allocation can provide the necessary incentives for users to behave such that the overall efficiency can be maximized. In this paper, we propose a dynamic pricing mechanism for the allocation of shared resources, and evaluate its performance. In contrast with several existing trading models, our scheme is designed to allocate a request with multiple resource types, such that the user does not have to aggregate different resource types manually. We formally prove the economic properties of our pricing scheme using the mechanism design framework. We perform both theoretical and simulation analysis to evaluate the economic and computational efficiency of the allocation and the scalability of the mechanism. Our simulations are validated against a prototype implementation on PlanetLab.
Citations
More filters
Journal ArticleDOI
TL;DR: This survey initially discusses all the relevant aspects motivating cloud interoperability, and categorizes and identifies possible cloud interoperable scenarios and architectures, and discusses future directions and trends toward the holistic approach in this regard.
Abstract: A brief review of the Internet history reveals the fact that the Internet evolved after the formation of primarily independent networks. Similarly, interconnected clouds, also called Inter-cloud, can be viewed as a natural evolution of cloud computing. Recent studies show the benefits in utilizing multiple clouds and present attempts for the realization of an Inter-cloud or federated cloud environment. However, cloud vendors have not taken into account cloud interoperability issues, and each cloud comes with its own solution and interfaces for services. This survey initially discusses all the relevant aspects motivating cloud interoperability. Furthermore, it categorizes and identifies possible cloud interoperability scenarios and architectures. The spectrum of challenges and obstacles that the Inter-cloud realization is faced with are covered, a taxonomy of them is provided, and fitting enablers that tackle each challenge are identified. All these aspects require a comprehensive review of the state of the art, including ongoing projects and studies in the area. We conclude by discussing future directions and trends toward the holistic approach in this regard.

405 citations


Cites background or result from "On Economic and Computational-Effic..."

  • ...Interoperability between different providers allows cloud customers to use the service across clouds to improve scalability and reliability [Mihailescu and Teo 2010c]....

    [...]

  • ...Market-based approaches for allocation of shared resources have proven their potential in computational systems [Mihailescu and Teo 2010a]....

    [...]

  • ...There is growing interest in the adoption of market-based approaches for allocation of shared resources in computational systems [Mihailescu and Teo 2010b]....

    [...]

  • ...They show that in their proposed dynamic scheme, the user welfare, the percentage of successful requests, and the percentage of allocated resources increase in comparison to the fixed pricing [Mihailescu and Teo 2010d]....

    [...]

Journal ArticleDOI
TL;DR: In this article, a peer offloading game among small-cell base stations (SBSs) is proposed to maximize the long-term system performance while keeping the energy consumption of SBSs below individual longterm constraints.
Abstract: The (ultra-)dense deployment of small-cell base stations (SBSs) endowed with cloud-like computing functionalities paves the way for pervasive mobile edge computing, enabling ultra-low latency and location-awareness for a variety of emerging mobile applications and the Internet of Things. To handle spatially uneven computation workloads in the network, cooperation among SBSs via workload peer offloading is essential to avoid large computation latency at overloaded SBSs and provide high quality of service to end users. However, performing effective peer offloading faces many unique challenges due to limited energy resources committed by self-interested SBS owners, uncertainties in the system dynamics, and co-provisioning of radio access and computing services. This paper develops a novel online SBS peer offloading framework, called online peer offloading (OPEN), by leveraging the Lyapunov technique, in order to maximize the long-term system performance while keeping the energy consumption of SBSs below individual long-term constraints. OPEN works online without requiring information about future system dynamics, yet provides provably near-optimal performance compared with the oracle solution that has the complete future information. In addition, this paper formulates a peer offloading game among SBSs and analyzes its equilibrium and efficiency loss in terms of the price of anarchy to thoroughly understand SBSs’ strategic behaviors, thereby enabling decentralized and autonomous peer offloading decision making. Extensive simulations are carried out and show that peer offloading among SBSs dramatically improves the edge computing performance.

246 citations

Journal ArticleDOI
01 Jul 2013
TL;DR: A revenue management framework from economics is adopted, and the revenue maximization problem with dynamic pricing as a stochastic dynamic program is formulated, and its optimality conditions are characterized, and important structural results are proved.
Abstract: In cloud computing, a provider leases its computing resources in the form of virtual machines to users, and a price is charged for the period they are used. Though static pricing is the dominant pricing strategy in today's market, intuitively price ought to be dynamically updated to improve revenue. The fundamental challenge is to design an optimal dynamic pricing policy, with the presence of stochastic demand and perishable resources, so that the expected long-term revenue is maximized. In this paper, we make three contributions in addressing this question. First, we conduct an empirical study of the spot price history of Amazon, and find that surprisingly, the spot price is unlikely to be set according to market demand. This has important implications on understanding the current market, and motivates us to develop and analyze market-driven dynamic pricing mechanisms. Second, we adopt a revenue management framework from economics, and formulate the revenue maximization problem with dynamic pricing as a stochastic dynamic program. We characterize its optimality conditions, and prove important structural results. Finally, we extend to consider a nonhomogeneous demand model.

232 citations

Journal ArticleDOI
TL;DR: The nature of noncooperative competition in an IAAS cloud market is characterized, with a goal of capturing how each IaaS cloud provider will select its optimal prices to compete with the others.
Abstract: As an increasing number of infrastructure-as-a-service (IaaS) cloud providers start to provide cloud computing services, they form a competition market to compete for users of these services. Due to different resource capacities and service workloads, users may observe different finishing times for their cloud computing tasks and experience different levels of service qualities as a result. To compete for cloud users, it is critically important for each cloud service provider to select an "optimal" price that best corresponds to their service qualities, yet remaining attractive to cloud users. To achieve this goal, the underlying rationale and characteristics in this competition market need to be better understood. In this paper, we present an in-depth game theoretic study of such a competition market with multiple competing IaaS cloud providers. We characterize the nature of noncooperative competition in an IaaS cloud market, with a goal of capturing how each IaaS cloud provider will select its optimal prices to compete with the others. Our analyses lead to sufficient conditions for the existence of a Nash equilibrium, and we characterize the equilibrium analytically in special cases. Based on our analyses, we propose iterative algorithms for IaaS cloud providers to compute equilibrium prices, which converge quickly in our study.

121 citations


Cites background from "On Economic and Computational-Effic..."

  • ...Teng et al. [8] and Mihailescu et al. [9] studied optimal pricing with an auction mechanism, in which users had budgetary and deadline constraints....

    [...]

Proceedings ArticleDOI
10 Jun 2012
TL;DR: Optimality conditions and structural results are obtained for the stochastic formulation of the infinite horizon dynamic pricing problem for an infrastructure cloud provider in the emerging cloud computing paradigm and these yield insights on the optimal pricing strategy.
Abstract: We study the infinite horizon dynamic pricing problem for an infrastructure cloud provider in the emerging cloud computing paradigm. The cloud provider, such as Amazon, provides computing capacity in the form of virtual instances and charges customers a time-varying price for the period they use the instances. The provider's problem is then to find an optimal pricing policy, in face of stochastic demand arrivals and departures, so that the average expected revenue is maximized in the long run. We adopt a revenue management framework to tackle the problem. Optimality conditions and structural results are obtained for our stochastic formulation, which yield insights on the optimal pricing strategy. Numerical results verify our analysis and reveal additional properties of optimal pricing policies for the infinite horizon case.

82 citations

References
More filters
01 Jan 2003
TL;DR: The BitTorrent file distribution system uses tit-fortat as a method of seeking pareto efficiency, which achieves a higher level of robustness and resource utilization than any currently known cooperative technique.
Abstract: The BitTorrent file distribution system uses tit-fortat as a method of seeking pareto efficiency. It achieves a higher level of robustness and resource utilization than any currently known cooperative technique. We explain what BitTorrent does, and how economic methods are used to achieve that goal. 1 What BitTorrent Does When a file is made available using HTTP, all upload cost is placed on the hosting machine. With BitTorrent, when multiple people are downloading the same file at the same time, they upload pieces of the file to each other. This redistributes the cost of upload to downloaders, (where it is often not even metered), thus making hosting a file with a potentially unlimited number of downloaders affordable. Researchers have attempted to find practical techniqes to do this before[3]. It has not been previously deployed on a large scale because the logistical and robustness problems are quite difficult. Simply figuring out which peers have what parts of the file and where they should be sent is difficult to do without incurring a huge overhead. In addition, real deployments experience very high churn rates. Peers rarely connect for more than a few hours, and frequently for only a few minutes [4]. Finally, there is a general problem of fairness [1]. The total download rate across all downloaders must, of mathematical necessity, be equal to the total upload rate. The strategy for allocating upload which seems most likely to make peers happy with their download rates is to make each peer’s download rate be proportional to their upload rate. In practice it’s very difficult to keep peer download rates from sometimes dropping to zero by chance, much less make upload and download rates be correlated. We will explain how BitTorrent solves all of these problems well. 1.1 BitTorrent Interface BitTorrent’s interface is almost the simplest possible. Users launch it by clicking on a hyperlink to the file they wish to download, and are given a standard “Save As” dialog, followed by a download progress dialog which is mostly notable for having an upload rate in addition to a download rate. This extreme ease of use has contributed greatly to BitTorrent’s adoption, and may even be more important than, although it certainly complements, the performance and cost redistribution features which are described in this paper.

2,985 citations


"On Economic and Computational-Effic..." refers background in this paper

  • ...In BitTorrent [7], rational users that behave selfishly and do not cooperate in sharing files are punished by other users....

    [...]

  • ...Thus, recent work in peer-to-peer networking [7, 16], grid or cluster computing [11], resource allocation [4,18], and others, use a form of incentives to manage rational users....

    [...]

Journal ArticleDOI
TL;DR: In this article, the seller's valuation and the buyer's valuation for a single object are assumed to be independent random variables, and each individual's valuation is unknown to the other.

2,435 citations


"On Economic and Computational-Effic..." refers background in this paper

  • ...From an economic perspective, the desirable properties for resource allocation are: individual rationality, incentive compatibility, budget balance and Pareto efficiency [12]....

    [...]

Proceedings ArticleDOI
08 Nov 2004
TL;DR: The goals of BOINC are described, the design issues that were confronted, and the solutions to these problems are described.
Abstract: BOINC (Berkeley Open Infrastructure for Network Computing) is a software system that makes it easy for scientists to create and operate public-resource computing projects. It supports diverse applications, including those with large storage or communication requirements. PC owners can participate in multiple BOINC projects, and can specify how their resources are allocated among these projects. We describe the goals of BOINC, the design issues that we confronted, and our solutions to these problems.

2,061 citations


Additional excerpts

  • ...For example, in BOINC [3], users donate their CPU cycles by running a software client which polls a server for new jobs....

    [...]

Proceedings ArticleDOI
14 May 2000
TL;DR: The proposed Nimrod/G grid-enabled resource management and scheduling system builds on the earlier work on Nimrod and follows a modular and component-based architecture enabling extensibility, portability, ease of development, and interoperability of independently developed components.
Abstract: The availability of powerful microprocessors and high-speed networks as commodity components has enabled high-performance computing on distributed systems (wide-area cluster computing). In this environment, as the resources are usually distributed geographically at various levels (department, enterprise or worldwide), there is a great challenge in integrating, coordinating and presenting them as a single resource to the user, thus forming a computational grid. Another challenge comes from the distributed ownership of resources, with each resource having its own access policy, cost and mechanism. The proposed Nimrod/G grid-enabled resource management and scheduling system builds on our earlier work on Nimrod (D. Abramson et al., 1994, 1995, 1997, 2000) and follows a modular and component-based architecture enabling extensibility, portability, ease of development, and interoperability of independently developed components. It uses the GUSTO (GlobUS TOolkit) services and can be easily extended to operate with any other emerging grid middleware services. It focuses on the management and scheduling of computations over dynamic resources scattered geographically across the Internet at department, enterprise or global levels, with particular emphasis on developing scheduling schemes based on the concept of computational economy for a real testbed, namely the Globus testbed (GUSTO).

965 citations

Proceedings ArticleDOI
17 Oct 2000
TL;DR: It is proved that the LP approach is an optimal allocation if and only if prices can be attached to single items in the auction, and suggests greedy and branch-andbound heuristics based on LP for other cases.
Abstract: When an auction of multiple items is performed, it is often desirable to allow bids on combinations of items, as opposed to only on single items. Such an auction is often called "combinatorial", and the exponential number of possible combinations results in computational intractability of many aspects regarding such an auction. This paper considers two of these aspects: the bidding language and the allocation algorithm. First we consider which kinds of bids on combinations are allowed and how, i.e. in what language, they are speci ed. The basic tradeo is the expressibility of the language versus its simplicity. We consider and formalize several bidding languages and compare their strengths. We prove exponential separations between the expressive power of di erent languages, and show that one language, \OR-bids with phantom items", can polynomially simulate the others. We then consider the problem of determining the best allocation { a problem known to be computationally intractable. We suggest an approach based on Linear Programming (LP) and motivate it. We prove that the LP approach nds an optimal allocation if and only if prices can be attached to single items in the auction. We pinpoint several classes of auctions where this is the case, and suggest greedy and branch-andbound heuristics based on LP for other cases.

609 citations


"On Economic and Computational-Effic..." refers background or result in this paper

  • ...However, an optimal allocation mechanism for multiple resource types such as combinatorial auctions requires a NP-complete algorithm [14]....

    [...]

  • ...Our results presented in Table 2 show that in different market conditions, dynamic pricing obtains better allocation results which result in increased economic efficiency....

    [...]