Distributed Computer Monitoring Based On CORBA

Distributed computer monitoring system is a kind of typical distributed computing environment. It has common features of distributed structure. At the same time, because of monitoring system's peculiarity it also has many problems needed to solve.At present time, there are some primary shortages in distributed computer monitoring system: Due to different communication protocols used in monitoring system, which brings on difficulties when we integrate system; The application software of monitoring system is poor in interoperability and reusability; The isomeric characters of net in computer monitoring make it hard to integrate system. In order to solve those problems, the author analyzes solution projects currently used in monitoring system: field-bus technology, OPC technology, middleware technology .By deeply analyzing and comparing these technologies, the author comes to conclusion: middleware is now the best solution project.Currently, CORBA and DCOM are main middleware's technologiesin the world. Both of them aim to make software based on object in distributed isomeric conditions has nicer nature on reuse, replanting and interoperability, so it can run well in variant hardware platform in the isomeric condition result from manifold OS, and then we can conveniently setup isomeric distributed application system. The author adopts distributed computer monitoring project based on CORBA, because CORBA 's capacity is better than DCOM 's in solving isomeric condition resulted from different OS. In order to validate this project's feasibility, the author sets up a simple model in experiment environment, and realizes it. Employing this model, application software's interoperability and reusability boost up greatly. At same time, this model also has virtue of enhancing system expansibility.This experiment's model adopts IDL language to define different equipments interoperable interface. This interface's IDL document is completely independent from different programming languages, OS, net protocol. Through this interface, application program can transparently call the equipment operation. Users may program different implementation based different field equipment protocols, and then fulfill the tasks by calling the interface. This experiment's model successfully solves the problems of integration in distributed computer monitoring in field equipment's variant communication protocols, and it also supplies a new way to system integration. This method not only has relatively highpracticability but also has good elicitation on the methods of system integration.