Dynamic Instance Provisioning Strategy In An LaaS Cloud Based On Openstack

In recent years, with the continuous development of Internet technology, Internet technology-based cloud computing applications are becoming more mature. Cloud computing can provide users with the service of inexpensive, scalable, distributed computing resource through the Internet. Computing resources are provided to the users in the form of virtual resources to meet varying user’s computing needs. Cloud service providers charge users for fees in accordance with the usage. The resource offered by cloud computing includes networks, servers, storage, memory, software, and other services to users.With the growing number of cloud computing users, the study of minimizing user fees under the premise of how to gurantee the user QoS requirements, has become an important research direction in cloud computing research field. Here, the QoS parameters include overload rate, delay rate, wait rate, resource utilization, etc.Recently, many open source cloud platform are emerging, such as OpenNebula, Eucalyptus, CloudStack, OpenStack, among which open source OpenStack cloud platform is the most widely used. OpenStack is an open source project, and it is a cloud platform for project management. It is a combination of several major open source components together to fulfill all the specific features of cloud platforms. The most notable feature of the OpenStack is that it can simplify the deployment process for cloud computing, and be easy to scale.This paper studies the dynamic resource allocation problem of the cloud platform. In the light of the current research basis, we deploy an OpenStack based cloud computing system, and design a dynamic resource allocation strategy to achieve the purpose of keeping the QoS parameter below a target threshold and reducing the total cost as much as possible.