CloudResearch云服务09Word下载.docx

1、D. BreitgandE. LevyA. GalisK. NaginI. LlorenteR. MonteroY. WolfsthalE. ElmrothJ. CaceresM. Ben-YehudaW. EmmerichF. GalanAbstractThe emerging cloud computing paradigm is rapidly gaining momentum as an alterna-tive to traditional IT. However, contemporary cloud computing offerings are mostly geared to

2、wards Web 2.0 style applications and only recently have begun to address the needs of enterprise-grade solutions, such as support for infrastructure-level SLAs.To address the challenges and deficiencies in the current state of the art, we propose a modular, extensible cloud architecture with intrins

3、ic support for Business Service Man-agement （BSM） and federation of clouds. Our goal is to facilitate an open, service-based, on-line economy, where resources and services are transparently provisioned and managed across clouds on an on-demand basis at competitive costs with high quality of service.

4、We present the vision driving the RESERVOIR project - an architecture that allows providers of cloud infrastructure to dynamically partner with each other to create a seem-ingly infinite pool of IT resources while fully preserving the autonomy of technological and business management decisions. To t

5、his end, RESERVOIR leverages and extends the ad-vantages of virtualization and embeds autonomous management into the infrastructure. At the same time, the RESERVOIR approach aims to achieve a very ambitious goal - creating a foundation for next-generation enterprise-grade cloud computing.1 Introduct

6、ionIn the Web 2.0 era, companies grow from inception to a massive scale at incredible rates. For example, MySpace acquired 20 million users in two years; YouTube reached the same number of users in just 16 months 1. However, to leverage this potential rate of growth, companies must properly address

7、critical business decisions related to their service delivery infrastructure.The emerging Cloud Computing paradigm 2, as exemplified by the Amazon Elastic Com-pute Cloud （EC2）, represents a promising conceptual foundation for hosting and deployment of web-based services while theoretically relieving

8、 service providers from the responsibility of provisioning the computational resources needed to support these services. Cloud computing offers multiple advantages: it allows individuals or companies with market domain expertise to build and run their Software as a Service （SaaS） company with minima

9、l effort in software devel-opment and without managing any hardware operations. This helps reduce software complexity and costs, expedite time-to-market, and enhance accessibility of consumers.With cloud computing, companies can lease infrastructure resources on-demand from a vir-tually unlimited po

10、ol. The “pay as you go” billing model applies charges for the actually usedThe research leading to these results is partially supported by the European Communitys Seventh Framework Programme （FP7/2001-2013） under grant agreement n 215605.1resources per unit time. This way, a business can optimize it

11、s IT investment and improve avail-ability and scalability.While cloud computing holds huge promise for the future of service computing, a number of inherent deficiencies in current offerings can be pointed out:Inherently limited scalability of single-provider clouds: Although most infrastructure clo

12、ud providers today claim infinite scalability, in reality it is reasonable to assume that even the largest players may start facing scalability problems as Cloud Computing usage rate increases. In the long term, scalability problems may be expected to aggravate as cloud providers serve an increasing

13、 number of on-line services, each accessed by massive amounts of global users at all times.Lack of interoperability among cloud providers: Contemporary cloud technologies have not been designed with interoperability in mind. This results in an inability to scale through business partnerships across

14、clouds providers. In addition, it prevents small and medium cloud infrastructure providers from entering the cloud provisioning market. Overall, this stifles competition and locks consumers to a single vendor.No built-in Business Service Management support: Business Service Management （BSM） is a man

15、agement strategy that allows businesses to align their IT management with their high level business goals. The key aspect of BSM is Service Level Agreement （SLA） manage-ment. Current cloud computing solutions are not designed to support the BSM practices that are well established in the daily manage

16、ment of the enterprise IT departments. As a result, enterprises looking at transforming their IT operations to cloud-based technologies face a non-incremental and potentially disruptive step.We argue that none of these problems, as well as other major problems such as security and availability, are

17、not remediable by retrofitting existing architectures. On the contrary, these issues should be addressed by proper design of a cloud computing architecture from basic principles. In this paper, we propose a reference model and architecture that systematically address some of those deficiencies and s

18、erve as a potential foundation for delivering IT services as utilities over the Internet.1.1 The RESERVOIR ApproachThe RESERVOIR vision is to enable on-demand delivery of IT services at competitive costs, without requiring a large capital investment in infrastructure. Our model is inspired by a stro

19、ng desire to liken the delivery of IT services to the delivery of common utilities. For example, a common scenario in the electric grid is for one facility to dynamically acquire electricity from a neighboring facility to meet a spike in demand. We deem that similarly to other industries, where no s

20、ingle provider serves all customers at all times, next-generation cloud computing infrastruc-ture should support a model where multiple independent providers can cooperate seamlessly to maximize their benefit.More specifically, we believe that to truly fulfill the promise of cloud computing, there s

21、hould be technological capabilities to federate disparate data centers, including those owned by separate organizations. Only through federation and interoperability can infrastructure providers take advantage of their aggregated capabilities to provide a seemingly infinite service computing utility

22、. Informally, we refer to the infrastructure that supports this paradigm as a federated cloud.An additional important advantage offered by the federated cloud approach is that it democ-ratizes the supply side of cloud computing and allows small and medium-sized businesses and new entrants to become

23、cloud providers. This encourages competition and innovation.2As discussed above, one of the limiting factors in current cloud computing offerings is the lack of support for BSM, specifically for business-aligned SLA management. While specific cloud computing solutions can be enhanced with some aspec

24、ts of BSM, the provisioning of complex services across a federated network of possibly disparate data centers is a difficult and yet unsolved problem. A service may be a composition of numerous distributed resources, including computing, storage, and network elements. Provisioning such a service con

25、sumes physical resources, but should not cause an SLA violation of any other running application with a probability larger than some predefined threshold. As SLAs serve as risk mitigation mechanisms, this threshold represents the risk that a cloud provider and the cloud customer are willing to accep

26、t.With BSM, applications are properly dimensioned, and their non-functional characteristics （e.g., performance, availability, security, etc.）, governed by SLAs, are ensured with optimal cost through continuous IT optimization. We argue that due to the immense scale envisioned by cloud computing, sup

27、port for BSM should be maximally automated and embedded into the cloud infrastructure.In the RESERVOIR model, each infrastructure provider is an autonomous business with its own business goals. A provider federates with other providers （i.e., other RESERVOIR sites） based on its own local preferences

28、. The IT management at a specific RESERVOIR site is fully autonomous and governed by policies that are aligned with the sites business goals. To opti-mize this alignment, once initially provisioned, resources composing a service may be moved to other RESERVOIR sites based on economical, performance,

29、 or availability considerations. Our research addresses those issues and seeks to minimize the barriers to delivering services as utilities with guaranteed levels of service and proper risk mitigation.Cloud computing is the latest incarnation of a general-purpose public computing utility. In recent

30、years we have seen the many efforts towards computing as a utility - from grid comput-ing 3, which made significant progress in the area of high performance scientific computing, to attempts at building enterprise-level utilities 4. However, none of these attempts have materi-alized into a general p

31、urpose compute utility accessible by anyone, at any time, from anywhere.What makes cloud computing different is that industry trends such as ubiquity of broad-band networks, fast penetration of virtualization technology for x86-based servers 5, and the adoption of Software as a Service 6 are finally

32、 creating an opportunity and a need for a global computing utility. The reluctance to use on-line services as a replacement for traditional software is lessening the success of companies such as proves that with the right set of security warranties and a competitive price, companies are willing to trust even their most valu-able data customer relations to an on-line service provider. At the same time, virtualization has made it possible to de