Olap Queries Based On P2p Distributed Storage Technology Research And Implementation

In recent years, OLAP (Online Analysis Processing, OLAP) has become a researchhotspot in the decision support field. Built on the data warehouse with massive data, itusually gives support to the Aggregate Query and multidimensional analysis. With thehelp of OLAP, the analysis staffs are easy to obtain all sorts of information viewthrough rapid, consistent, multifaceted and interactive access. OLAP is the primarymeans to acquire decision support. In the application process, the user may have a highrequirement on the response speed of OLAP query, so how to improve the efficiency ofOLAP query and analysis has been a key issue in the data warehouse application, thenmany experts and scholars have done a lot of research such as SOL query optimization,materialized view selection, the improvement of the clustering algorithm, the data cubestorage, query rewrite, data compression, data cache and so on.With the time passes by and the data amount becomes larger and larger, thetraditional OLAP technology structure need to be improved in order to better meet theneeds of the user decision analysis. P2P (Peer-to-Peer) technology put forward a newidea for the network application and lead in the concept of peer entity. The advantagesare as follows: quick access, avoid single point failure and eliminate the bottleneck ofthe central node, etc. It gains many researchers’ attention gradually, and the use of P2Ptechnology for OLAP decision analysis has become a hot research. In the P2Penvironment, we can use data cube among multiple OLAP network nodes to completethe decision analysis. Because it can balance the OLAP server load and improve theefficiency of OLAP query and analysis.This thesis is to establish a kind of model in the P2P environment, that is all theOLAP nodes working together to complete the decision analysis. And this model isrealized mainly through the following aspects:(1) According to the traits of P2Penvironment, it makes use of the cube’s semantics dimension hierarchy chain features tofinish the Data Cube store and puts forward a P2P routing algorithm so as to ensure theOLAP node quickly find the data source needed by decision analysis from the P2P network.(2)After completing the data source storage in the P2P environment, the datareplication, as one of the key technologies for data management, will inevitably causedata consistency maintenance issues. In order to improve the data availability andsystem performance, this thesis puts forward a way of maintaining data consistency.(3)The thesis completes a design of OLAP query in the P2P network. That is toconstruct the OLAP network model in P2P environment to realize the dynamicaladdition and exit of OLAP nodes, and propose a kind of modified analysis algorithm toachieve the data share and cooperation of multi-node data cube.