本文選題：緩存 + 分布式��； 參考：《電子科技大學》2012年碩士論文

【摘要】：隨著國內(nèi)外互聯(lián)網(wǎng)的快速發(fā)展，，分布式緩存技術(shù)也得到了廣泛的應(yīng)用。幾乎所有的互聯(lián)網(wǎng)公司都會使用到分布式緩存技術(shù)來提高系統(tǒng)的吞吐率，降低對數(shù)據(jù)庫的訪問頻率，提高系統(tǒng)的快速響應(yīng)能力�；ヂ�(lián)網(wǎng)中的應(yīng)用系統(tǒng)都是屬于查詢密集型的，也就是數(shù)據(jù)的查詢操作數(shù)遠遠大于其插入的操作數(shù)。所以分布式緩存技術(shù)在互聯(lián)網(wǎng)中的應(yīng)用基本上都是用于提高系統(tǒng)的查詢效率的。但是雖然分布式緩存技術(shù)在查詢密集型系統(tǒng)中得到了廣泛的應(yīng)用，但是并不意味著分布式緩存技術(shù)只能應(yīng)用在查詢密集型的系統(tǒng)中。該論文的研究方向就是把分布式緩存技術(shù)應(yīng)用到插入密集型的系統(tǒng)中，以提高系統(tǒng)的最大吞吐率。 本文依托于國家信息產(chǎn)業(yè)部電子發(fā)展基金項目“高性能、高可靠入侵防御系統(tǒng)研發(fā)與產(chǎn)業(yè)化”。針對項目中對系統(tǒng)吞吐量的需求，在該系統(tǒng)中引入分布式緩存技術(shù)。并重點研究了在插入密集型的系統(tǒng)中引入分布式緩存技術(shù)后的系統(tǒng)架構(gòu)問題、分布式緩存的管理以及分布式緩存下實現(xiàn)數(shù)據(jù)冗余的方法，同時使用該論文中的研究成果最終實現(xiàn)了一個分布式緩存系統(tǒng)。 本文的主要研究成果包括： 1)通過對插入密集型系統(tǒng)進行分析和研究，完成了在該系統(tǒng)中引入分布式緩存技術(shù)后的系統(tǒng)架構(gòu)的設(shè)計。 2)在研究和分析一致性hash算法和虛擬節(jié)點技術(shù)在分布式緩存系統(tǒng)中的應(yīng)用以后，提出了基于一致性hash的數(shù)據(jù)冗余技術(shù)。將數(shù)據(jù)按照一定的算法保存在不同的緩存服務(wù)上，降低因系統(tǒng)故障造成數(shù)據(jù)丟失的風險。 3)在研究和分析常用的緩存分布式管理技術(shù)以后，提出了集中管理、本地詢問的緩存管理技術(shù)。并詳細的描述了管理數(shù)據(jù)初始化、管理數(shù)據(jù)修改的流程。 4)通過在系統(tǒng)的實現(xiàn)中使用非阻塞通信技術(shù)、基于JSON的序列化以及concurrent同步技術(shù)，實現(xiàn)了一個高速、高并發(fā)的分布式緩存系統(tǒng)。
[Abstract]:With the rapid development of the Internet at home and abroad, distributed caching technology has been widely used. Almost all Internet companies will use distributed cache technology to improve the throughput of the system, reduce the frequency of access to the database, and improve the system's ability to respond quickly. The application systems in the Internet are query intensive, that is, the query operands of data are much larger than the operands inserted. So the application of distributed cache technology in the Internet is basically used to improve the query efficiency of the system. But although distributed cache technology has been widely used in query intensive systems, it does not mean that distributed cache technology can only be applied in query intensive systems. The research direction of this paper is to apply distributed cache technology to insert intensive system to improve the maximum throughput of the system. This paper is based on the National Ministry of Information Industry Electronic Development Fund project "High performance, high reliability intrusion prevention system R & D and industrialization." To meet the requirement of system throughput in the project, distributed cache technology is introduced into the system. This paper also focuses on the system architecture after introducing the distributed cache technology into the inserted intensive system, the management of the distributed cache and the method of realizing the data redundancy under the distributed cache. At the same time, a distributed cache system is implemented using the research results in this paper. The main research results of this paper are as follows: 1) through the analysis and research of the insert intensive system, The design of the system architecture after introducing the distributed cache technology into the system is completed. 2) after the research and analysis of the application of the consistent hash algorithm and the virtual node technology in the distributed cache system, Data redundancy technology based on consistent hash is proposed. In order to reduce the risk of data loss caused by system failure, the data is stored on different cache services according to certain algorithms. 3) after researching and analyzing the commonly used distributed cache management technology, centralized management is proposed. Local query cache management technology. And described in detail the management data initialization, management data modification process. 4) through the use of non-blocking communication technology in the implementation of the system, JSON-based serialization and concurrent synchronization technology, the realization of a high-speed, High concurrency distributed cache system.
【學位授予單位】：電子科技大學
【學位級別】：碩士
【學位授予年份】：2012
【分類號】：TP333

【引證文獻】

相關(guān)碩士學位論文 前1條

1 姚小勇;基于Memcached的中國移動CRM系統(tǒng)性能分析及優(yōu)化[D];中山大學;2013年

本文編號：2068139