Font Size: a A A

The Research And Analysis Of Dynamic Resource Scheduling Mechanism In PaaS Environment

Posted on:2016-02-17Degree:MasterType:Thesis
Country:ChinaCandidate:T G ZhangFull Text:PDF
GTID:2308330482960434Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the development of information technology and the rise of cloud computing, PaaS(Platform as a Service) platform draws wide attentions in recent years because of its practical environment and resources. It provides easy ways for the management of the cloud computing resources, which reduces the business operation costs significantly. How to realize the flexible platform resource configuration and improve resource utilization is a valuable research topic. The research and analysis of dynamic resource scheduling mechanism is the focus of study.In this thesis, it introduces the specific design and implementation of a dynamic resource scheduling mechanism in PaaS environment, which used the C programming language and Ruby/Rails programming framework.Firstly, this thesis gives an overview about the research and analysis of dynamic resource scheduling mechanism in PaaS environment. It introduces the mathematical modeling and analysis of the platform system to solve the resource monitoring and resource scheduling problems. Secondly, it lays out the structure of the new-designed resource scheduling system and carries out a detailed introduction to resources discovery module, scheduling policy module, scheduling action module, as well as the internal process and interface between the modules. It introduces the specific design and implementation of the core algorithm in the system based on the previous design. Apart from this, this thesis gives an detailed description of the scheduling algorithm and the realization of the resource migration mechanism in chapter 5. In addition, a trend analysis method for monitoring metrics is proposed, which includes the regression model and the exponential smoothing prediction model to predict the values of monitoring metrics in future time.At last, it designs a functional tests based on open source Cloud Foundry platform. Finally, the contribution, deficiency, and future development of the research are discussed.
Keywords/Search Tags:cloud computing, platform as a service, forecasting analysis, monitoring metrics
PDF Full Text Request
Related items