A method for realizing automatic scaling and load balancing of a distributed database system based on kvm cloud platform

Abstract

The invention relates to a method for achieving automatic extensible load balancing based on a distributed database system of a KVM cloud platform. The method specifically comprises the steps that each database server node monitors CPU usage amount itself, an unloading requirement is sent to an MON process when the mean value of peak values of the CPU usage amounts is greater than or equal to a set threshold value upper limit of a Timer time period in the Timer time period, and a load increase requirement is sent to the MON process when the mean value of the peak values of the CPU usage amounts is smaller than or equal to a set threshold value lower limit of the Timer time period. The method establishes a high-performance distributed relational database, can continuously provide services in a cluster and perform automatic extension under the situation that the data consistency is kept to achieve automatic load balancing, and the possibility of the database downtime due to over-high load is directly reduced.

Description

technical field [0001] The invention relates to a method for realizing automatic scaling and load balancing of a distributed database system based on a KVM cloud platform, and belongs to the technical field of cloud computing. Background technique [0002] KVM, short for Kernel-based Virtual Machine, is an open source system virtualization module that has been integrated in all major distributions of Linux since Linux 2.6.20. It uses Linux's own scheduler for management, so compared to Xen, its core source code is very small. KVM has now become one of the mainstream VMMs in academia. [0003] Because of its advantages in cost and performance and its unique on-demand distribution, cloud computing platforms are widely replacing traditional IT service server architectures, from traditional information service platforms to distributed computing, big data mining, machine learning and other fields. While saving a lot of hardware costs and labor costs, it provides a basic platfor...

AI Technical Summary

Problems solved by technology

[0004] Common systems such as e-commerce systems and transaction systems based on relational data cannot be effectively replaced by other types of databases in a short period of time due to the special performance requirements of the database. The carrying capacity of the database also puts forward higher requirements. The performance of traditional single-point relational databases is subject to the development of electronic hardware performance, and it cannot be greatly improved in a short period of time. Therefore, distributed relational databases have also become widespread concern. Object

Images