Reseach On Marching-on In-time Scheme And The Fast Algorithm Of Time Domain Integral Equation

In similar to petroleum engineering type enterprise operations of informatization construction in the process, the existence of large different branches of a single node information operations work become very difficult, in the lands in operations staff cost is too high, such as difficult to secure.We through the analysis of a geographically distributed computer room environment and research, network monitor and control system of early warning information and message queue, and the combination of expert system, to provide first-hand information for operational service desk, generated the repair order for service desk provides a source of alarm.This article mainly analyzes the kei igawa drilling geophysical exploration, in a distributed network, the company focus on distributed deployment operations monitoring, early warning message transmission, the respect such as expert system alarm message identification are studied, put forward a set of monitoring early warning system in the distributed environment.Thoughts and methods of the research is based on domain driven design, with Nagios to monitor software, asynchronous message queue, expert system as the main technology of research and development, design method, using UML2 as adhere to the agile development process in the implementation process, to frame is suitable for the operational early warning system of distributed computer room environment as the basic goal, and the implementation of a distributed operations to meet the requirements of ITIL implementation of early warning system.This project was conducted in accordance with the requirements for kei igawa the company’s operations, for such as geophysical prospecting company, international engineering are around the room, such as operational level is not consistent with the situation to carry on the research and design, according to the operational monitoring of high real time capability and have prior notice requirements, considering the monitoring early warning system usually has a lot of low level alarm and repeat the same alarm, false alarm of normal maintenance downtime, etc., a large number of information must be through the expert system for preassigned, only in the transmission of the news of important to the operational platform;Exceptions may occur during transmission network, to guarantee the accuracy of the information to transfer to the operational platform, can use the asynchronous message queue technology;Due to the distributed environment, the warning message usually come from different monitoring system, will have a reasonable matching the information and get the relatively accurate fault location.Warning system mainly provide service failure to declare, is before the fault handling process, reduce the alarm noise is system must consider the key factors.A implementation of the system is through the deployment of open source Nagios system, combining the RabbitMQ asynchronous message queue, use Ruby on Rails plug-ins + expert system research and development, the main purpose is to validate its feasibility.The warning message output in line with the Restful Service standard format, for the future integration into the whole operational framework provides support.The biggest support system for 2 nodes, nodes can be extended to more in the future.Through the research and the implementation, this paper expounds on the distributed computer network monitoring early warning, the decentralized control technology, can let a small amount of technical experts focus on one to solve the problem quickly, so as to enhance operational efficiency.