Monitoring System In Virtual Machine Management Platform

With the upgrading of computer hardware performance, how to take full use and maximum effectiveness of computing resources has become a concern of universities and enterprises. The development of virtualization technology provides a new solution to the issue.In this paper, based on the research content of multi-system resources virtualization and in conjunction with the actual high-performance cluster computing environment, we propose VNIX, a multi virtual machine management platform in cluster computing environment. The architecture of VNIX can divide into three hierarchies. Basic services layer dispersed in computing nodes of cluster which contain management of virtual machines in single computing nodes, collection of dynamic system information and persistent. Core services layer deploys on the management node of cluster which manage all virtual machines in computing nodes in global. Core services layer mainly contains fast deployment of virtual machines according to application requirements and dynamic load balance. User interface layer creates a friendly graphical interface by which administrators can deploy many virtual machines in a short time and have a direct viewing of status of physical machines and virtual machines.Monitor and control is the most fundamental function of virtual machine management. When an administrator get down to some management work on a large-scale system, first and foremost he must have a clear understanding of the up-to-the-minute system conditions. Only after understanding what is going on, can an administrator determine exactly what to do with the system. Besides, to fulfill most tasks deploy and load balance also has to resort to monitor and control.The function test and the performance test show the success design and efficient performance of the VNIX Monitoring system. Comparing to the similar systems, the network bandwidth overhead is largely decreased by approximately 83.6%.