Memory Caching Mechanism In Trade Database

It has really become a hot topic to increase the efficiency in a database system with large amount of data and high intensity of incoming requests.This research is also based in a large Stock trade system using relation database which may be stuck by some disk-writing operation like updates or commitments within these times. Our goal is to pick up some principles of memory and real-time database and build up a memory (main memory) based mechanism for this system in order to increase the efficiency of data processing and operations for customers, without corrupting the construction or throwing out bulks of money to buy a real-time database.This research is worked with a real project of Kingstar Group, and confirm to a basic flow of Software Engineering. In Analysis Step, we got requirements and the target of data processing, and analyzed many things such as which table suits for memory caching; In Design Step, we build up a mainframe of this mechanism and the structure of memory table with its operating functions. Meanwhile we create some index links using RBTrees, and a rule for concurrence, sharing, monitoring and synchronization. In the efficiency and intensity tests afterward, we trusted decuple and 3 times' rises of speed in writing and the whole processes respectively. At last we had considered about the problems in maintenance in order to draw a long live to this system.