1、Cloud Computing Architecture 云計(jì)算體系結(jié)構(gòu),Tsu-Yang Wu,2,SPI軟件-平臺(tái)-基礎(chǔ)設(shè)施 model,Generally, the classification scheme for cloud computing has been coined the SPI model Software-Platform-Infrastructure This acronym首字母縮寫(xiě) represents the three major services provided through the cloud SaaS, PaaS, and IaaS,3,SPI E

2、volution,To understand how the SPI framework evolved, its helpful to place it in context with the development of Internet Service Providers互聯(lián)網(wǎng)服務(wù)供應(yīng)商 (ISPs). 1.0 As ISPs originally began to provide Internet services, dial-up modem service for homes and organizations proliferated, making the Internet a

3、 commercial commodity.,4,SPI Evolution,2.0 During a period of merging and consolidation, ISPs began offering other services, such as e-mail and off-site data storage. 3.0 The increasing demand for infrastructure to host their customers applications and data led to the creation of data centers Multip

4、le customers could centralize their servers, storage, and communications systems on the ISPs premises.,5,SPI Evolution,4.0 The commoditization商品化 of collocation facilities led to the development of application service providers應(yīng)用服務(wù)提供商 (ASPs). 5.0 The ASP model eventually evolved into cloud computing

5、 New delivery models, such as the SPI framework, with its SaaS, PaaS, and IaaS Service models, and various deployment models, such as private, community, public, and hybrid cloud models.,6,SPI evolution through virtualization虛擬化,Data center evolution through basic virtualization into a full SPI fram

6、ework,7,SPI Framework Service,IaaS 基礎(chǔ)設(shè)施即服務(wù) Shared Internet infrastructure, such as servers and storage服務(wù)器和存儲(chǔ)器 Amazon EC2 and S3, Sun Microsystems Cloud Services, Terremark, Dropbox, 阿里云 PaaS 平臺(tái)即服務(wù) Application platform that provides developers with quick deployment部署 Google App Engine, Microsoft Azur

7、e,8,SPI Framework Service,SaaS 軟件即服務(wù) Stateless cloud-enabled multiple-instance applications on a pay-per-use pricing model Apples MobileMe & Google Docs Security VS. Extensibility 安全和可擴(kuò)展 trade-offs權(quán)衡 SaaS - Least extensibility最小可擴(kuò)展 and greatest amount of security responsibility taken on by the cloud

8、 provider云提供者負(fù)責(zé)安全責(zé)任 IaaS - Greatest extensibility最大可擴(kuò)展 and least amount of security responsibility taken on by the cloud provider PaaS - Middle and security features that must be leveraged by the customer,9,Software as a Service (SaaS),NIST 美國(guó)國(guó)家標(biāo)準(zhǔn)協(xié)會(huì) The capability provided to the consumer is to use

9、the providers applications running on a cloud infrastructure The applications are accessible from various client devices through a thin client interface such as a web browser (e.g., web-based e-mail). The consumer does not manage or control the underlying cloud infrastructure, including network, ser

10、vers, operating systems, storage, or even individual application capabilities,10,Software as a Service (SaaS),The SaaS customer rents the usage of the software using an operational expense model a pay-per-use or subscription agreement訂購(gòu)合同 The pay-per-use按次計(jì)費(fèi) licensing model is also known as on-deman

11、d licensing請(qǐng)求式許可 As some applications delivered交付 through the SaaS model are billed on a metered usage計(jì)量收費(fèi)使用 and time-period basis, rather than paying the common upfront costs of traditional licensing,11,Platform as a Service (PaaS),PaaS is similar to SaaS, but the service is an entire application d

12、evelopment environment應(yīng)用部署環(huán)境, not just the use of an application. accessible via a Web browser Using the concept of PaaS, software developers can build Web applications without having to install the software building tools on their own computer,12,Platform as a Service (PaaS),NIST The consumer is to

13、 deploy onto the cloud infrastructure consumer-created消費(fèi)者自創(chuàng)內(nèi)容 or acquired applications created using programming languages and tools supported by the provider. The consumer does not manage or control the underlying cloud infrastructure.消費(fèi)者不能管理和控制底層云基礎(chǔ)設(shè)施 but has control over the deployed applications

14、 and possibly application hosting environment configurations.,13,Platform as a Service (PaaS),Software development platform to be considered a true PaaS Baseline monitoring of application usage should be used to effect platform process improvement. To provide seamless integration無(wú)縫集成 with other clou

15、d resources Web-based databases基于網(wǎng)絡(luò)的數(shù)據(jù)庫(kù) and other Web-based infrastructure components and services基于網(wǎng)絡(luò)的基礎(chǔ)設(shè)施及服務(wù) Security安全, privacy隱私, and reliability可靠 must be maintained as a basic service. The development platform must be browser-based開(kāi)發(fā)平臺(tái)必須基于瀏覽器.,14,Infrastructure as a Service (IaaS),NIST The con

16、sumer is to provision processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications.消費(fèi)者可以部署和運(yùn)行任意的軟件 The consumer does not manage or control the underlying cloud infrastruc

17、ture(消費(fèi)者不能管理和控制底層云基礎(chǔ)設(shè)施) but has control over operating systems, storage, deployed applications,15,Infrastructure as a Service (IaaS),IaaS can deliver either basic or complex storage capabilities as a service over the Internet. IaaS可以實(shí)現(xiàn)基本或復(fù)雜的存儲(chǔ)容量 This enables the pooling and sharing of hardware resou

18、rces 合并既共享硬件資源 such as servers, storage (drives or sans), and perimeter devices (firewalls, routers).,16,Infrastructure as a Service (IaaS),IaaS example,17,Cloud Deployment Models,Cloud deployment models are often referred to as external or internal clouds. Each of these models, must share the funda

19、mental tenets of cloud computing Each deployment model employs Internet-connected devices (每個(gè)模型使用聯(lián)網(wǎng)設(shè)備) Each model provides for dynamic scaling of virtual resources (每個(gè)模型提供動(dòng)態(tài)擴(kuò)展的虛擬資源) Users of each model commonly dont have control over the technology being used,18,Cloud Deployment Models,Public cloud

20、公有云 The cloud infrastructure is made available to the general public or a large industry group and is owned by an organization selling cloud services. Private cloud 私有云 The cloud infrastructure is operated solely for an organization.(私有云被組織單獨(dú)擁有) It may be managed by the organization or a third party

21、.(可能被機(jī)構(gòu)或第三方管理),19,Cloud Deployment Models,Community cloud 社會(huì)云 The cloud infrastructure is shared by several organizations and supports a specific community that has shared concerns (云基礎(chǔ)設(shè)施被許多組織共享) mission, security requirements, policy, and compliance considerations It may be managed by the organizat

22、ions or a third party,20,Cloud Deployment Models,Hybrid cloud 混合云 The cloud infrastructure is a composition of two or more clouds (兩種以上云合成的云基礎(chǔ)設(shè)施) private, community, or public Remain unique entities Are bound together by standardized or proprietary technology that enables data and application portab

23、ility e.g., cloud bursting for load-balancing between clouds(云之間的負(fù)載平衡),21,Public Clouds 公有云,It is generally open for use by the general public. The general public is defined in this case as either individual users or corporations個(gè)人用戶(hù)或公司. The public cloud infrastructure used is owned by a cloud servi

24、ces vendor organization公有云的擁有者是云服務(wù)提供商 Amazon Web Services, Google App Engine, S, and Microsoft Windows Azure The cloud is operated and managed at a data center owned by a service vendor hosts multiple clients and uses dynamic provisioning動(dòng)態(tài)配置,22,Public Clouds,Using a public cloud can provide immedia

25、te cost savings直接成本節(jié)約 to an organization. Shared infrastructure, remote hosting遠(yuǎn)程主機(jī), and dynamic licensing and provisioning are strong enticements誘惑 for a company.,23,Private Clouds 私有云,Using virtualization, some companies are building private cloud computing environments使用虛擬化技術(shù)，一些公司建立私有云計(jì)算環(huán)境 intend

26、ed to be used only by their employees or designated partners.公司雇員或指定合作伙伴使用私有云 Offer the benefits of public cloud computing Enabling the organization to retain greater control over the data and process.組織可以最大程度上控制數(shù)據(jù)和處理過(guò)程,24,Private Clouds,Some features 私有云的特征 Private cloud deployment appears to be sp

27、earheaded primarily by larger organizations and governmental agencies, rather than small companies or end users. 私有云最開(kāi)始只是在大的組織或政府機(jī)構(gòu) A private cloud requires much of the same infrastructure outlay that traditional IT infrastructure requires, so an institution would need to be fairly large to get bene

28、fits from the private cloud model A larger company would also have the need for full control and access to the internal firewalls內(nèi)部防火墻 and virtualized environment虛擬化環(huán)境 required to build the private cloud.,25,Private Clouds,Private VS Public cloud Infrastructure associated with a private cloud is com

29、monly dedicated to a single enterprise and is not shared with any other enterprise.私有云只應(yīng)用與單獨(dú)的企業(yè) Security is considered to be tighter in a private cloud deployment than it is in a public cloud.安全問(wèn)題在私有云中更嚴(yán)格,26,Hybrid Clouds 混合云,Example An organization deploying noncritical software applications in the

30、 public cloud, while keeping critical or sensitive apps in a private cloud. 組織部署非緊急軟件應(yīng)用在公有云，緊急/敏感的軟件應(yīng)用在私有云。,27,Alternative Deployment Models,The Linthicum Model - David Linthicum He sees 10 major categories, or patterns, of cloud computing technology Storage as a Service Database as a Service Inform

31、ation as a Service Process as a Service Application as a Service (SaaS) Platform as a Service (PaaS) Integration as a Service Security as a Service Management/Governance as a Service Testing as a Service,28,Expected Benefits 預(yù)期效益,Major benefits 主要效益 Means to move from operating in a capital expendit

32、ure environment to an operational expenditure environment Ability to rapidly deploy innovative business and research applications in a cost-effective manner通過(guò)經(jīng)濟(jì)的方式迅速部署創(chuàng)新企業(yè)和研究應(yīng)用 Use of virtualization to detach business services from the underlying execution infrastructure通過(guò)虛擬化技術(shù)分離商業(yè)服務(wù)和底層執(zhí)行基礎(chǔ)設(shè)施 Disaster recovery and business continuity capabilities are intrinsic in the cloud paradigm災(zāi)難恢復(fù)和業(yè)務(wù)連續(xù)性能力在云范式中是本質(zhì)的,29,Expected Benefits預(yù)期效益,Ability of the cloud provider to apply s