Font Size: a A A

Research On Multi-Tenant Service Composition Approach Based On Service Granularity Space

Posted on:2015-01-19Degree:MasterType:Thesis
Country:ChinaCandidate:H H CaiFull Text:PDF
GTID:2268330431455495Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
In recent years, cloud computing has received wide attention from academia and industry. As a common delivery model in cloud computing, SaaS model is becoming increasingly popular. With the increasing of user’s individual and diverse requirements, multi-tenancy has been the main delivery model for SaaS applications in future. Meanwhile, in order to adapt to the complex application development, application components had become modularization and fine-grained. Thus, new applications can be built through assembling those components quickly and agilely. So service composition for multi-tenant is the key to build application flexibly. However, with the development of cloud computing, tenant’individual and diverse requirement become more intense. Traditional service composition approaches cannot adapt to the service composition for multi-tenant in cloud computing, which makes multi-tenant service composition tend to be more challenging. The challenges reflect in the following four aspects:Business requirements are always diverse, some are abstract and some are specific. Requirements show characteristic of multi-granularity. The characteristic of multi-tenant in SaaS mode makes almost every tenant has special requirements for the application. Because of tenants’individual requirements, they can customize services to build applications they need. It indicates characteristic of individuation which varies a lot. Then there is another correlation caused by customization among customized services and original services. Services always frequently evolve from one version to another version due to business requirements changed.In order to solve the challenges above, this paper proposes a service granularity space model, which provides a semantic basis for multi-tenant service composition. The service granularity space supports the characteristic of hierarchy, inheritance, evolved correlation and versioning, effectively responding to the challenge of service composition for cloud computing SaaS applications. On the one hand, service granularity space makes it possible to change disorder services into the hierarchy and ordered clustering services, and on the other hand, makes it easy and efficient to develop multi-tenant SaaS applications. Besides, the approach proposed in the paper is applied to the social security service system development based on SaaS and we design and implement prototype system.The main contributions made by this thesis include:1) Through research on the characteristic of multi-granularity, massive customization, correlation caused by customization and frequent evolution of multi-tenant service composition, this paper proposes the service granularity space that supports the characteristic of hierarchy, inheritance, evolved correlation and versioning. It can provide as a semantic basis for multi-tenant service composition and effectively respond to complexity of service composition for cloud computing SaaS application.2) Through research on the service clustering and correlation relationship techniques, this paper proposes the approach of building the service granularity space. It turns huge services disorder to order and can construct multi-tenant application quickly.3) Through research on the characteristic of hierarchy of service granularity space proposed, this paper proposes the algorithm of multi-tenant service composition based on service granularity space. It guarantees service composition plans of different granularity and satisfies the QoS requirements, which realizes the diverse requirements.4) Because of the great service space, it consumes too much time to select service for each abstract service. It has seriously affected the performance of multi-tenant service composition. So this paper proposes the algorithm of pruning service space based on skyline computing, which makes the performance of service composition greatly improved.The approach proposed in the paper is applied to the social security service system development based on SaaS, in which we introduce the implementation of individuation, verify the building of service granularity space, the approach of service space pruning and the approach of multi-tenant service composition based on the service granularity space.
Keywords/Search Tags:cloud computing, multi-tenant, service composition, service granularityspace
PDF Full Text Request
Related items