An application-aware virtual machine scheduling method for load balancing in IaaS environment

Abstract

The invention discloses an application-aware virtual machine scheduling method oriented to load balance in an IaaS environment, comprising the following steps: step 1, configuring automatically related files for a virtual machine introspection library LibVMI and a memory analysis framework Volatility on the host by an OpenStack; the application-aware virtual machine scheduling method comprises thefollowing steps: 2, acquiring application perceptual data and storing that data in a database; 3, selecting a virtual machine to be migrated on each overload host machine to obtain a virtual machineset to be migrated; Step 4: Picking an appropriate host for each virtual machine to migrate and migrate it using real-time migration technology. Compared with the traditional method, the invention effectively reduces the number of SLA breaches under the condition of frequent load changes, and realizes better load balancing on CPU, memory, bandwidth and other resources, and has good adaptability and practical value.

Description

【Technical field】 [0001] The invention belongs to the field of cloud computing, and in particular relates to an application-aware virtual machine scheduling method oriented to load balancing in an IaaS environment. 【Background technique】 [0002] As a new type of network computing model, cloud computing provides users with various virtual IT resources and application services through the network in an on-demand and easy-to-expand manner, and supports users to access from any location and using multiple terminals. According to the different services it provides, cloud computing can be divided into three types: Infrastructure as a Service (Infrastructure as a Service, IaaS), Platform as a Service (Platform as a Service, PaaS) and Software as a Service (Software as a Service, SaaS). ). IaaS services mainly provide users with the IT infrastructure needed for business in the form of virtual machines. These virtual machines are deployed in the data centers of cloud service provid...

AI Technical Summary

Problems solved by technology

However, for highly dynamic workloads, forecast-based load balancing methods are not very reliable because it is difficult to accurately predict resource requirements and usage.

[0008] To sum up, the existing work has improved the load balance between the hosts in the data center to a certain extent, but most of the work in the process of virtual machine scheduling only regards the virtual machine as a black box, from the outside Observing its resource requirements without knowing anything about its internal situation does not solve the load balancing problem of virtual machines well

Images