The Research Of Distributed Cache System Based On Cloud Computing

The rapid development of the Internet and the rapid growth of the global numberof Internet users make the amount of data network services grow exponentially. Cloudcomputing as a new service-oriented computing model, has the great advantages ofhighly scalable resources, powerful computing and storage capacity. Therefore, a triplebasic framework based on cloud computing IaaS, Paas, and SaaS is built and deployedas a distributed caching system.First of all, this paper introduces the background and significance of the research,and the present conditions in this field, Subsequent analysis and describes thecharacteristics, architecture, typical applications and key technologies of cloudcomputing. The formation of the main contents in this article is below:1. The DCache distributed caching system is built based on the basis of Dynamosystem framework, and analyzed combined with the current limitations of the traditionalrelational database needs. Distributed system needs to be supported by the three needs:high concurrent read and write, high efficiency of mass data storage and access, highscalability and availability of the database.2. DCache distributed caching system data cache layer is designed andimplemented. The system data cache technology uses the memory allocation and datadeletion in the memcached protocol, and a variety of storage modes are designed tosatisfy various application scenarios. Based on this design and detail, the function of thecache layer and data layer framework is introduced in detail.3. BDB embedded database is used as persistent storage of the system, and thepersistent model based on the order of the memory index (IFS) is used to solve theproblem of inefficient mass data storage and read BDB database are facing with. IFSmodel can make the order of writing instead random writing, and make creation,modification, deleting become the supplements of files. Index of the data is stored inmemory, always pointing to the latest data, so that the massive data storage and readspeed are greatly improved.4. In this paper, the mass of high-performance distributed caching system DCachedata is designed, writed and tested scenarios,in JAVA and the environment.netrespectively, based on its current application scene. And the second response times(IOPS) and bandwidth (throughput) these two indicators are used for assessing the ability of data access, and the success rate for assessing the reliability of the system.The final test results prove that the advantages of concurrency and data processingin the DCache system has meet the expected needs in the face of massive data.