Network Operations Management System

With the advanced development of computer application, the requirement of job management system based on cluster technology which can manage the resources effectively is becoming very urgent in recent years. To meet these needs, the Network Job Management System (NJMS) is designed. NJMS system is one kind of system software based on OS and its main goal is to reinforce the job management function of OS, use the resource of system effectively, balance the network load of system and improve the performance of the whole system.The system architecture, load balance mechanism, job network scheduling strategy and the performance optimizing process of NJMS system are introduced in this thesis. The main results are as follows.1. The load balance mechanism research of NJMS system. Deeply discussed the workgroup cluster architecture and the load balance realization of NJMS system based on this architecture. Analyzed the deficiency of workgroup cluster. Put forward the concept of network cluster and the load balance mode based on machine network cluster and gave the implement of this mode.2. The research of job network scheduling strategy. Analyzed job scheduling strategy in NJMS system. Deeply discussed some job scheduling arithmetic of NJMS system based on single machine and multi-machine. Put forward a batch request scheduling arithmetic based on multi-machine, moreover, accomplished the testing and evaluating of this arithmetic.3. The optimization of NJMS system. Deeply researched the main three problems, including multi-time zone, multi-code and memory leak. Given the solving methods of these problems.NJMS system is a long-term international cooperant project of NWPU software engineering center. Now the first phase task has been finished. NJMS system has been passed checking and accepting by some foreign experts. Moreover, many commercial products, including Chinese, Japanese and English version were released. The research of the second phase is going to start and the main work is focused on the research based on grid computing and the continual optimization of the system performance.