A dynamic memory management model on Xen virtual machine

Xen visualization has been widely used in cloud computing, data center and cyber physical systems. Because memory can not be used by time-sharing as other devices, how to efficiently allocate memory has become the bottleneck of performance improvement on Xen. In this paper, we propose a dynamic memory management model on Xen virtual machine with the goal of availability growth, which includes a fast memory adjustment strategy with memory recycling, memory increasing, and memory adjustment. Experimental results show that, in order to satisfy the memory requirements virtual machines need, this memory management model can recycle a large amount of free memory in virtual machines rapidly; in the aspect of increasing memory, using exponential growth mode can also satisfy high memory requirements in a short time, which has improved availability of the virtual machine requiring memory urgently; and in the case of shortage of memory resources, the model can also give priority to meet the demand of virtual machine requiring memory urgently, which has avoided availability falling due to shortage of memory resources.