Font Size: a A A

The Research And Design Of Distributed Data Cache Mechanism

Posted on:2014-04-27Degree:MasterType:Thesis
Country:ChinaCandidate:X LiuFull Text:PDF
GTID:2268330425959885Subject:Software engineering
Abstract/Summary:PDF Full Text Request
As the age of big data is coming, distributed application syste m has become thema instream of develop ment. As a kind of important means for improving applicationperformance, distributed data cache has beco me an ind ispensable link in today’sdistributed application deve lopment, it also has been paid wide attention and stud iedby enterprises. Currently many open source product obvious had shortcomings interms of data redundancy backup, fa ilures and sca lability etc, and commerc ia l cachesystem is the high cost and opaque, whic h leads to the open source products need todo some extens io ns to meet app licatio n require ments. Therefore how to research anddesign a distrib uted data cache fra mework used to solve these proble ms is becomingvery urgent.Through studying current existing distributed data cache technolo gy, especia llyin the aspect of data redundanc y backup, fa ilures and scalab ility etc, the thesis haspointed out the shortages and interna l factors caused these defic ienc ies. On this basis,the thesis put forward a k ind of generic, low cost, and a scalab le distributed datacache fra mework. The fra mework is designed based on the existing servicecompone nts, whic h inte grates the advantages of each service compone nt, also ensuresscalab ility and ava ilability of the who le fra mework. At the same time, dataredundancy backup of the fra mework inc lud ing fa ilo ver and load ba lanc ing featurescan be imple me nted, it a lso expands system cache replace ment strategy, themecha nis m of distributed transaction commit and network I/O model etc. In order toensure the availab ility of the syste m, every compone nt has realized redundancymecha nis m, whic h make it own stro ng ab ility of fa ult to lerance. In add ition, syste mconcurrent ability can be constantly improved by horizontal scaling.Fina lly setting up a simulative test environme nt, it is verified that the mechanis mof its various functions can be run automatica lly and correctly. Performance andconcurre ncy ab ility have reached the design goal a nd the fra mework can comp letelymeet the needs of practical application. At last it is proved to have certain practica lvalue and engineering significance.
Keywords/Search Tags:Distributed data cache, data redundanc y backup, fa ilover, load bala nc ing, availability, distributed transaction
PDF Full Text Request
Related items