Research And Application Of Performance Optimization Of Health Monitoring Big Data Platform Based On Hbase

Through the mobile Internet,intelligent equipment and Internet technology,people can track the current physical health indicators,their movement status,diet and other lifestyle habits.Collecting these data can help people dig up more valuable medical information,so as to give something back to the more the broad masses of the people.The development of the Internet technology,however,has not kept up with the pace of the data increases.There are still a lot of problems in large scale data storage,management and real-time query.At the same time,there is no uniform standard for health monitoring data.This will undoubtedly affect the development of health monitoring of big data.In order to solve the above problems,this paper draws on some of the existing health monitoring data processing scheme,researchs on the principle of Hbase on the big data processing,as well as its advantages and disadvantages,by which we puts forward build the health monitoring big data platform by using Hbase and bulid a unified health monitoring data model by using <the urban and rural residents' health records basic data set>.Hbase is still in its stage of development at present,which is not able to perform complex data query,unable to meet the demand of the development of the business.Therefore,this paper studies the performance optimization of Hbase,and presents some query optimization algorithms on Hbase query,and puts forward the load prediction algorithm on the Hbase overall load,so the main work is as follows:(1)Study the hadoop,hbase related technologies,construction of big data storage platform;(2)Study <the urban and rural residents' health records basic data sets> to build a unified health data storage model;(3)Research on big data real-time processing architecture,build appropriate health monitoring of large data storage architecture.Put forward the use of hidden column family build Hbase secondary indexes,according to bloom filter and Hash join algorithm Hbase table join optimization algorithm is proposed,Hbase statistical query optimization scheme is proposed.On the whole load of Hbase,puts forward the presplit Hash algorithm and load forecasting algorithm based on exponential smoothing,aimed at to improve the performance of Hbase to meet the business requirements.(4)Finally,the experiment system was set up to complete the performance test and the experiment was analyzed.The experimental results show that the proposed to Hbase optimization scheme and algorithm can more effectively from the aspect of writing and query data improve Hbase performance.