Data Integration Environment Based On The Log View Surveillance Methods To Achieve

At information explosive times, to integrate needed information timely and effectively from diverse information sources become more and more important. Data integration technique is an efficient way to solve this problem.Data integration system consists of monitoring intelligent agent and integrator. This paper is studying and implementing monitoring way so as to the information sources send initiatively and continuously new information with uniform format to its subscriber according to the predefined demands and the data change on itself.At first, this paper puts forward notion of view monitor based on traditional table monitor. The function of view monitor is to detect the change of view and propagate it to integrator. The main meliorations include three ways: 1. The monitored object is changed from all tables to the tables about view. In this way, monitor avoids unnecessary data changes. 2. The information that monitoring agent propagate to integrator is changed from table delta to view delta. It reduces traffic and decreases the burden of integrator. 3. Monitoring agent propagate information to integrator according to the rule of subscriber. In this way, propagation and integration become more efficient.Secondly, we fulfill monitoring way based on log. The primary work includes three aspects :Firstly, we adopt log analyzer. The function of log analyzer is to monitor log contents periodically, distill interrelated log records and change them into log analysis data flow.Secondly, view delta calculation, whose purpose is to calculate view delta based on log analysis data flow, adopts deferred view monitoring method.Thirdly, data communication, which include log analysis data flow propagation and view delta propagation, adopts Socket communication based on TCP/IP protocol.