Design And Implementation Of Load-Balanced Query System Based On TERADATA Warehouse

The competition amount of the enterprises is increasingly fierce with economicdevelopment. Meanwhile, the ways and channels of enterprises to get information areincreasing, they have to analyze the data needed from the huge and disorder ones, and maybeit is a chance to catch up or beyond the completion. It is always one of research during thedata warehouse building, and how to quickly query the data from it, how to standardize thequeries, and how to design the system to be a load balancing and security to dealing with theuses queries, also the queries are still under evolving.The Query System (QS) is one tool for information query from the Teradata warehouse,and load balancing system based on Teradata warehouse. The QS based on specific norms anddefinitions, to control the frequent and disorder of the query request. It’s not only forstandardized of the query, but also for the rational allocation of the human resources and otherresources, and for the monitoring of query request. With the technique of load balancing, it isensure the high availability of the QS, and the work efficiency.During the study and the implementation process, which amount of information systemsand most of them need to interact and to use the database, then gives a cross-platform,language-independent, loosely coupled programming model, which is based on the proceduralframework of Web Service. And compare the two design language, including Java and.Net,also their development tools and theirs performance, and consideration of the independent forthe programmer and the software development cycle, drawn that based on design architectureof.Net Web Service technology has advantages. The main contributions of this thesis are asfollows:1. For the network load balancing, especially for the web system, there show thescheme using Network Load Balancing architecture based on window server. Simplydeploying applications to them, having the online hot backup, ensuring the stability of thesystem. Then it puts forward the core function of QS–the query engine. Through the queryengine control model, including cache control model and counter control model, fully use of hardware resources, it is one solution to the problem, and then solved the system load querybalancing.2. The thesis shows a method called universal naming convention. When the userrequests a query cache data, they do not need to worry about the location of the data in thenetwork. The QS only need to ensure that the cache data in unison amount the server nodes,only submit query requests, it will be hinted and then return the cached results; combined withsoftware designed to solve the problem of data synchronization.3. Combination of these key technologies, it carries on the design and implementationof the load balancing query system based on the Teradata warehouse, and has two testprograms, namely whether to adopt the network load balancing technology, then comparisonand analysis the test results, finally obtained the result: the query system based on.Net WebService technology, and network load balancing provided by window server, has highavailability, high scalability, and load balancing capabilities and other features.