Research On Subscribe & Publish Mechanism Of Log Module In Real-time Database

The real-time database system is used to gather, process, preserve and inquire the large amount of industrial scene data. Besides this, it also implements error analysis and performance optimization by making use of the important events collected, the information of operations, running status and alarms. Thus it's able to monitor the industrial production process. As the unceasing development of information construction in industrial enterprises, reducing the production cost and optimizing production process have been the core to improve the enterprises' competitive power. In order to meet this need, at present most real-time database products have implemented the unique log systems. According to different ways of information acquisition, log systems are divided into two kinds: systems of inquiring monitoring information periodically and systems of sending monitoring information actively. The former one is a passive information acquisition way; but the latter one driven by the monitoring information satisfies the actual monitoring process better. Subscribe&publish model is used in systems of sending monitoring information actively. At present, log systems implemented in this way are not many, iHistorian product promoted by GE Fanuc Corporation has implemented the subject-based subscribe&publish model.The thesis has implemented content-based subscribe&publish model, the main task is as follows: the architecture design, functional design and API design of log system in real-time database are completed; subscribe&publish is as a core of the log system and the system has also implemented collection, subscribe&publish, preservation and confirmation of alarms; the system provides other modules in real-time database with C APIs in the form of dynamic link library.The log system referred to in this thesis is a module of real-time database. The system has implemented content-based model. Therefore, it's better than iHistorian in the expressiveness about event and subscription condition. The system has also provided the more detailed inquiry mechanism by which users gain the statistical information of event and alarm conveniently. In general, the log system is much closer to the demand of real-time database user.