Research And Implementation Of Cloud Resources Optimization Management

Cloud computing is an emerging computing business models and services. It makes computing resources concentrated as a virtual machine of large server clusters provides users(tenants) with a dynamic, highly reliable, and scalable on-demand services and enables users to access to computing power, storage space and information services. Therefore, the configuration and optimization of cloud computing resources will be a key factor affecting the performance of cloud computing.This paper focuses on the user demand for resources, user classification and calculation type and considers a cloud computing environment resources to deploy and configure load balancing and energy-optimized operation. The main research work includes the following:(1) In a private cloud application scenario for the deployment of resources, from the user’s perspective, this paper presents the concept of key users and ordinary users, the allocation of resources among the different user type configuration scheme; From the perspective of dynamic allocation of resources, the comprehensive consideration of task type, degree of resource utilization, and can accommodate, we propose a virtual machine placement initialization and dynamic updates placement strategy.(2) For some cloud computing environment deployment ignore the influence of load balancing and task type problem, we consider the resource constraints, type of task and resource equilibrium factor, propose load balancing virtual machine deployment algorithm based on Ant Colony Optimization. Finally, we do experiment in the open source cloud simulation platform CloudSim and compare with the other two algorithms to prove the virtual machine deployment algorithm is verified.(3) For some cloud computing environment deployment ignore the overall utilization, the increase of energy consumption and higher costs problem, we propose a multi-objective optimization of low power consumption and load balancing algorithm, take the load balancing of multiple solutions as a candidate solution set and get a further selection of low power consumption. Finally, we do experiment in the open source cloud simulation platform CloudSim, the experimental results show that the virtual machine deployment algorithm ensures low energy consumption based on load balancing.