SuperComputer Operations Management System’s Analysis And Design

The rapid development of super computing power system has made a strong management needs of the operation support. With the continuous development of our computing center and hardware resources increasing gradually, operation and maintenance work has been concerned day by day. Every super computing center has its specific operation pattern while they still have unified operational management standards. Establishing a safe, scientific and informational ized operation management system will help to utilize the natural resource of our super computing center. It also will help to develop the whole operation quality and expand professional skill. Besides, it can also ensure machine keep running and improve the degree of satisfaction of client.The principal business of super computing center includes technical assistance, machine leasing, scheme making and scientific research project cooperation. Our paper mainly focuses on some aspects of maintenance of equipment, knowledge system, comprehensive information searching and operation control and so on. We analyze business operations and machine operations of our super computing center and point out the designing scheme including system module partition, database design and implementation method of each module based on the demand analysis. The function module is consisted of event tracking and analysis, electronic parameter, information management, the dynamic knowledge base, bordereaux, on-site inspection maintenance and service assurance. In order to study complex and massive data processing, events (fault) tracking, operation process, knowledge base and so on in the super computer environment, we establish related optimization model algorithm, and implement operating maintenance with the property of high quality, safety, specifications, high efficiency and low cost. Eventually we have exploited a suit of operation management system for National Supercomputing Center (Ji’nan center).The system takes B/S architecture design. This way can greatly help to alleviate the client load, reduce the cost and the workload of system maintenance and upgrading, and reduce the development of our system and overall cost of the use. This system has the advantage of simple deployment, powerful, complete security system and so on.Systematic development adopts the basic idea of JSP dynamic Webpage technology and modular design. The system has good interactivity, scalability and can support multi-platforms. Our system design view intelligence, efficience, and ease of use as the principle. By the establishment of the system, we can make the operation and management of effective planning, organization, management, processing, timely tracking and feedback to the event and fault, and the knowledge base also provides a public technology and knowledge sharing platform for maintenance and management staffs.