Design And Implementation Of Micro-Blogging Website

In2010, the Web2.0website and social network service (SNS) have raised the public’sattention in China. The requirements of high concurrent read and write in database fromwebsite, the requirements of the efficient storage and visitation of mass data as well asscalability and high availability for database are increasing. MicroBlog is one of therepresentor. The current mainstream MicroBlog is using the distributed practice Mysqldatabase to provide data support. But at the present time, there are still some problems cannotbe solved, that is how to realize the goal such as easily for system to expand, low delay, highavailability, distributed monitoring and decentralization. In order to solve these problems,based on the comprehensive analysis of several non-relational database system, this passagedecided to use a non-relational database Cassandra as a solution for data storage. Build up aMVC framework through the Spring, Struts and Hector, in order to built a MicroBlog systemwhich allow a large amount of data read and write at the same time. The use of databaseCassandra for the data storage, because it has a flexible mode (especially in large scaledeployment), real data scalability, multiple data center identification characteristics; and italso provides the range query, list data structure and distributed for writing operation,. Thismakes it more competitive than other traditional databases in function. There will be aconstruction of7column group in Cassandra in this paper. By adding, deleting and updatingthe column group, the realization of the basic functions will come true， such as release, takecomments, forwarding, and focus on users. For system which only use one singleCassandra database server, there is no great difference of capability for its data processingand the use of relational database. By the limitation of experimental conditions, it is unableto provide specific performance data for reference. For no constraints to a non-relationaldatabase, data organization is allows to be relatively scattered, ability of data processing isstrong and allows to operate analyze data mining. Hope that as a beginning, the future ofnon-relational database can have wide development space.