The Memory Global Optimization Technology Approach Based On Virtualized Cloud Computing Environments

With the rapid development of Internet technology, the trend which is the growth of information services and more and more people relying on it has been unstoppable. However, the internet service providers have to face the challenges of delivering better service at lower cost for their customers. Nowadays cloud computing gradually becomes the necessary choice by bussiness users, for the reason of cloud computing's pay-as-you-go model and its' super elevation cost performance. Based on virtualization technology, cloud computing center can deliver at-scalability and on-demand services in an abstract fashion from underlying components, with the traits of easier manageability, simple maintenance and high utilization. However, the combination of cloud computing and virtual ization technology introduces a new pattern on resource allocation and utilization. Resource-on-demand based on virtualization technology can promote the resource utilization, increase the Quality of Service (QoS) and even decrease the Total Cost of Ownership (TCO).Furthermore, the resource boundaries of physical servers unfortunately restrain the global optimization of resources in cloud computing. And the existing reaserches are generally based on the platform of the cloud.Some researchers optimize the scheduling algorithms of memory resources among each virtual machine to increase rate of overall memory resource utilization.Others improve methodologies from aspects of architecture, address mapping, page swapping and idle pages recovering. As a result, those methods can improve the rates of memory resource utilization well, but it also have flaws. So the virtual resources in cloud computing still need to be reallocated, scheduled or memory mapped, otherwise, it can indeed increase the burden of virtual machine and reduce rate of memory resources utilization.In this paper, we present some methodologies and technologies to improve the framework of exiting global optimization, and add the strategy of minimum value of boundary when virtual machines have idle memories. And at that time, it need to compare the space of memory with minimum value of boundary, then map the rest of memory space into the global idle memory pool. Based on the above framework, the virtual machine's memory resources optimization research can be divided into two situations:lower resource utilization and higher resource utilization. At lower situation, virtual machine will first compare minimum value of boundary and put part of idle memory to global free memory pool. At the other higher situation, it will start global schedule through global free memory pool. Doing so will not only reduce the times of exchanges between virtual machine and global free memory pool, but also reduce the times of exchanges among virtual machines. As a result, average rate of memory resources utilization will be greatly improved.This paper presents the experiment analysis for the virtual memory resource global optimization methdologies. To sum up, our research is a beneficial exploration of memory resource cooperation and sharing technology in cloud virtualization resource. The experiment results show that the improving framework and algorithms are well done for the cloud resource configuration. Not only does it promote the utilization rate of cloud resource, but also it increases the reliability and efficiency on cloud platform.