基于Hadoop和SSAS的IT服务级别管理系统设计与实现

【作者】 张宇；

【导师】 戴开宇；

【作者基本信息】 复旦大学 ， 软件工程（专业学位）， 2013， 硕士

【摘要】 IT服务级别管理是IT服务管理中居于核心地位的管理过程。企业实施IT服务级别管理需要有力的软件工具支持才能有效应对不断变化的业务环境和IT环境。针对目前缺乏合适的IT服务级别管理软件工具的问题,本文介绍了一个基于Hadoop平台和SQL Server分析服务(SQL Server Analysis Services, SSAS)构建的IT服务级别管理系统。文章首先分析了IT服务级别管理的概念和意义,结合目前企业实施IT服务级别管理的实际需求,指出现有软件工具的主要问题在于底层数据库管理系统的处理能力不足以及缺乏对面向业务的服务级别管理的有效支持。然后使用领域建模技术对IT服务级别管理系统的需求做了详细的分析,提出了一个能够应对需求快速变化的面向业务的IT服务级别管理领域模型。在此基础上,设计了基于Hadoop/Hive, SQL Server/SSAS和SharePoint等技术的系统架构,并详细讨论了数据仓库、对象模型、管理门户和报表／仪表盘四个部分的设计与实现。论文的意义主要体现在两个方面：其一,把基于Linux操作系统的Hadoop与基于Windows操作系统的SSAS结合起来,构建了一个具有相当程度自动化特征的IT服务级别管理系统,在满足大数据处理需求的同时,有效地支持了企业对面向业务的IT服务的服务级别管理过程。其二,利用SSAS的在线分析处理功能实现了服务级别计算引擎,并与SharePoint, Excel等工具集成,在满足各类企业报表需求的同时大大降低了开发工作量,对此,文中讨论了完整的解决方案及实现的关键细节。更多还原

【Abstract】 IT Service Level Management is the central management process of IT Service Management. A modern enterprise needs powerful support from software tools to be adapted to the continuously changing business environment and IT environment, if it plans to implement IT Service Level Management. The thesis described an IT Service Level Management system built on top of Hadoop platform and SQL Server Analysis Services (SSAS) as a solution to the problem that enterprises lack appropriate software tools for IT Service Level Management.The thesis first analyzed the concept and significance of IT Service Level Management, and pointed out that the main problems of existing software tools are the lack of underlying database management system with big data processing capabilities and the lack of effective support for business-oriented IT Service Level Management. Then it analyzed the requirements of an IT Service Level Management system in detail and proposed a domain model for business-oriented service level management by using domain modeling techniques. Based on the model, the thesis designed the system architecture which leverages the technologies like Hadoop/Hive, SQL Server/SSAS and SharePoint, and discussed the design and implementation details of data warehouse, object model, management portal and reports/dashboards.The thesis introduced two innovations. First, it built up an IT Service Level Management system with considerable automation features by using Linux-based Hadoop and Windows-based SSAS. The combination of Hadoop and SSAS makes it possible to meet the requirements of processing big data and meanwhile effectively support the implementation of business-oriented IT Service Level Management. Second, it developed a calculation engine by leveraging the Online Analytical Processing (OLAP) features of SSAS and integrated the engine with Microsoft business intelligence tools like SharePoint and Excel. The solution greatly reduced the development effort and meanwhile kept the ability to meet various enterprise reporting requirements. The thesis discussed the whole solution and key implementation details.更多还原