Díaz-Sánchez, Felipe
Cloud brokering: New value-added services and pricing policies / Felipe Díaz-Sánchez - Bogotá: Universidad Santo Tomás, 2016.
vii, 115 páginas; ilustraciones, gráficas.
Incluye referencias bibliográficas (páginas 99-108)
ISBN 978-958-631-961-4
1. Almacenamiento virtual (computación) 2. Comercio electrónico
3. Computación en la nube I. Universidad Santo Tomás (Colombia).
CDD 004.67820688 CO-BoUST
© Felipe Díaz-Sánchez
© Universidad Santo Tomás
Ediciones USTA
Carrera 9 No.51-11
Edificio Luis J. Torres, sótano 1
Bogotá, D. C., Colombia
Teléfono: (+571) 5878797 ext. 2991
editorial@usantotomas.edu.co
http://ediciones.usta.edu.co
Directora editorial
Matilde Salazar Ospina
Coordinación de libros
Karen Grisales Velosa
Asistente editorial
Andrés Felipe Andrade
Diagramación
Julián E. Morales Ortega
Diseño de cubierta
Kilka Diseño Gráfico
Hecho el depósito que establece la ley
ISBN: 978-958-631-961-4
Primera edición: 2016
Todos los derechos reservados. Se prohíbe la reproducción total o parcial de esta obra, por cualquier medio, sin la autorización previa por escrito de los titulares.
Contents
About this book
Cloud brokering
Part I. Value-added services in cloud brokering
Cloud performance and placement in cloud brokering
Introduction
Cloud performance evaluation
Motivations and challenges
Studies related to cloud providers performance evaluation
Cloud Virtual Machine (VM) characterization
Placement in cloud brokering
Non-functional requirements-based placement
Application aware placement
Conclusion
Towards a figure of merit for cloud performance
Introduction
Performance evaluation
Evaluation methodology
Experimental setup
Provisioning time
Computation benchmarks performance
Memory benchmarks performance
Storage benchmarks performance
Variability
Figure of merit of VM cloud performance
Mean and radar plot as figures of merit
Simple figure of merit
Figure of merit based on Analytic Hierarchy Process
Case study: CPU-intensive application
Summary
An exact approach for optimizing placement in cloud brokering
Introduction
Goal programming
An exact approach for the Placement problem
Parameters
Variables
Goal
Constraints
Case study: Online trading platform
Part II. A new pricing model in cloud brokering
The Pay-as-you-book pricing model
Introduction
Pricing models in cloud computing
Advance Reservations
Advance Reservation specified by cloud providers
Advance Reservation specified by end-users
Pay-as-you-book
Initial scheduling of Advance Reservations
Pricing and rewarding end-users
Resource allocation policies
Case Study: A Virtual cloud Provider maximizing revenues through the Pay-as-you-book pricing model
Experimental setup
Results and analysis
Summary
Conclusion and future works
Appendix A. Studies related to cloud providers performance
Appendix B. Cloud performance evaluation: details and extended results
Related issues to the performance evaluation
VM configurations
Benchmark duration
Performance-price correlation with a simple figure of merit of Cloud performance
Correlation among VM sizes from different cloud providers
Correlation among different VM sizes from a single cloud provider
Bibliography
List of Figures
List of Tables
About this book
Service arbitrage enables advanced services in cloud brokering by taking advantage of two or more cloud provider offers. This allows cloud brokers to simplify the vast number of offers by categorizing the features and benefits of each cloud provider in order to match consumer needs with an ideal set of cloud providers. In the first part of this book entitled "Value-added services in cloud brokering", a comprehensive, state of the art study has been carried out on cloud performance evaluations and placement in cloud brokering (Chapter 1). Then, a method is proposed to calculate cloud performance through a single figure of merit based on the mapping of the physical features of a Virtual Machine (VM) to their respective performance capacities (Chapter 2). Finally, an exact placement approach for optimizing the distribution of cloud infrastructure across multiple providers is proposed (Chapter 3). Parameters such as price, VM configuration, VM performance, network latency and availability are considered for that purpose.
Nowadays, pay-as-you-go and reserved pricing dominate the way consumers acquire cloud resources from legacy cloud providers at the infrastructure level. However, the introduction of cloud brokers may induce the commoditization of cloud infrastructures. Facing such an evolution, new pricing models are necessary to capture potential consumers or untapped market segments. The second part of this book entitled "A new pricing model for cloud brokering" focuses on the design of a pricing model for cloud brokering, called pay-as-you-book (Chapter 4). Pay-as-you-book is based on two types of information. The first type consists of the forecast of users’ job requests. The second one consists of the ability of cloud brokers to take advantage of such advanced reservations. With this aim in view, a study comparing three resource allocation policies under pay-as-you-book is carried out. The aim of this book is to contribute to the design of new value-added services and pricing models for cloud brokering. The majority of the investigations and original results presented in this manuscript have been achieved and obtained in the context of the CompatibleOne [1] research project supported by the French Ministry of Industry. Its objective was to demonstrate the feasibility of a cloud brokering intermediation platform integrating and adapting the various software solutions proposed by the industrial and academic partners of the project. This platform provides a single point for service consumption in order to avoid vendor lock-in. This book has three objectives:
The contribution of this book can be itemized as the following:
This book consists of an overview of the following conference publications:
Cloud brokering
The role of cloud Brokers in the near future of cloud computing has been identified by Gartner as a major market trend: “By 2015, cloud Brokers will represent the single largest category of growth in cloud computing, moving from a sub-$1 billion market in 2010 to a composite market counted in the hundreds of billions of dollars.” [2]. This prediction seems to be reinforced by the amount of funding raised by some cloud brokering companies: Rightscale US$47.3m in three rounds1, 6fusion US$10m in two rounds, cloud Cruiser US$7.6m in two rounds, Zimory Systems US$7.2m in two rounds and Gravitant US$3.7m in one round [3]. One of the main reasons, behind this high economic expectation, is the highly heterogeneous current cloud market constituted by many cloud providers. Each cloud provider exhibits different interfaces, pricing models and value-added services. To help the end-user cope with such a fragmented ecosystem, cloud brokers have emerged as an intermediary third-party that provides unified self-service access to multiple cloud providers. Thus, by being a single point for service consumption, cloud brokers provide interoperability and portability of applications across multiple cloud providers. Besides this inherent role, current cloud brokers provide other value-added services to cloud consumers, such as the following: Advanced management by using tools beyond the stacks offered by cloud providers (e.g. consolidated billing, infrastructure monitoring, disaster recovery, SLA enforcement), elasticity management in order to automatically scale up or down infrastructure resources based on the workload and service arbitrage with the aim of taking advantage of two or more cloud provider offerings (e.g. cost optimization). These services can be overlayed, enabling new cloud computing scenarios such as cloud bursting or cloud marketplaces (Figure 1). These new scenarios may be beneficial for both end-users and cloud providers. In the case of cloud bursting, end-users have the possibility of extending their computing facilities by moving the development of applications or the non-mission-critical applications to public clouds. In the case of a cloud marketplace scenario, end-users have access to multiple cloud providers through a single interface, while cloud providers may sell spare infrastructure capacity. Cloud brokers are expected to drive creation of
Figure 1: Evolution and dependency of value-added services in cloud brokering
value through advanced value-added services enabling new cloud computing scenarios. The price of cloud computing resources varies around 20% between cloud providers, while the difference in performance between cloud providers remains unknown or less studied [3]. Due to the fact that cloud brokers are able to deploy a workload in any cloud provider, the measurement of performance for cloud providers and the placement of cloud resources based on a cost-performance relationship may be supported by cloud brokers in future value-added services. Moreover, the commoditization of infrastructure resources will increase the cloud adoption rate by simplifying the purchase of cloud computing resources. As cloud computing resources are traded like any other commodity (e.g. wheat, oil, iron) the currently fragmented cloud market will flatten. This will open the door to new pricing models in which cloud brokers not only act as intermediaries but also as liquidity providers, negotiating volume discounts from cloud providers and guaranteeing resource availability to end-users.
____________________
1A funding round is a practice by which a company raises money to fund operations, expansion, an acquisition, or some other business purpose.
Esta obra se editó en Ediciones USTA, Departamento Editorial de la Universidad Santo Tomás.
2016
To Leta and Days.
Part I
Value-added services in cloud brokering
Chapter 1
Cloud performance and placement in cloud brokering
Introduction
As the number of cloud computing services increases, so does the interest of consumers to be able to compare these services in order to choose those best adapted to their needs. This chapter focuses on the performance issues related to cloud provider evaluation and on the role of cloud brokers in the automatic optimization of resource allocation across multiple cloud providers. This chapter is structured as follows. Section "cloud performance evaluation". presents a survey of the current studies related to cloud performance evaluation. The motivation for and challenges behind the evaluation of cloud provider performance is also detailed in this section. Section "Placement in cloud brokering" describes the current methods used to allocate resources in cloud brokering. The studies are classified into two categories: placement based on non-functional requirements and application-aware placement.
Cloud performance evaluation
Motivations and challenges