Design And Implementation Of The Monitoring System To The Topology Information Collection

The monitoring system is a very important tool in the study of computer architecture and performance. It provides the quantity reference for the improvement and the new architecture by getting the performance, payload and efficiency data of the computer hardware and software. It is very meaningful to design the large scale distributed system monitoring and management based on topology information collection system.Both the study of the application, story, measuring technology of the topology information and the study of the monitoring system and management of the large scale distributed system is reviewed. The responding architecture in which the distributed data are collected and concentrated to deal is designed.The model of Probe and Agent is used in collecting the distributed topology information and transmitting the concentrated data on the Linux. The multithreaded probes collect the payload information and task status information and the payload information is gotten from proc, the virtual file system and the task status information is gotten from the shared memory. It has the characteristic of load payload, independence, favorable extendibility and strategy.The model of the manager and Agent and the SNMP are used for transmitting the data between the monitoring station and the monitoring center. The structure of the payload information and system task status information is designed based on SNMP.At last, in order to keep control command and strategy real time, integrity, secrecy, RC4 is adopted to encrypt the transferred data, md5 is adopted to hash the transferred data and heart beat is adopted to detect the system failure.