Research And Development Of Information Service Monitoring System

With the increase in the collections of library electronic resources, the popularity of information management, library information systems have been expanding, the network structure has become increasingly complex. Introduction of information service monitoring system to improve the availability of hardware, software and network services, health status and cause of the failure analysis report is to improve service levels, improve business efficiencies of digital library and an important means of increasing user satisfaction.Library information service monitoring system is an independent software for the library switches, servers, specific process and imitating access services and databases to be monitored.System can monitor all kinds of software, hardware, services, operation of the network and load conditions ((including servers, switches, databases, imitating access services, specific processes and MAC), to identify problems in time and help staff to detect system failures and eliminate waiting for the user to report situation; simulate user access behavior, to detect abnormal access to electronic resources, improve the reader’s satisfaction; achieve automatic alarm mechanism, through the mail or SMS promptly notify the system administrator, to prevent the problem from occurring, and gradually establish automatic fault warning mechanism; to improve the work efficiency of the network administrator, timely troubleshooting.System can gets switch run state information and interface various state information, and generated state figure, administrator can view state trend; system can gets Server run state information and each disk use situation; system can detection to server specific process run state, and judgment database connected situation and by simulation user into information service system gets service State; system can view network panoramic figure and network topology figure; system can view MAC address and IP address information. In addition, the system can configure detected items, customize the alert threshold, text and mail in a timely manner to the network administrator.The context model-based anomaly detection and analysis of data collected relatively long period(six months) draw the distribution of the various system resources, as well as changes in cycle. Anomaly detection results can provide a reference for the set threshold.