Method and system for realizing fault tolerance of virtual machine

Abstract

The invention provides a method and a system for realizing fault tolerance of a virtual machine. The method comprises the steps of: redundantly operating the virtual machine where an application program is located in a plurality of servers, selecting one server as a master virtual machine, using the rest servers as slave virtual machines, wherein the virtualization configurations of the master virtual machine and the slave virtual machines are the same, and the master virtual machine and the slave virtual machines are redundant backups for fault switching; and enabling the master virtual machine and the slave virtual machines to carry out redundancy processing on the user request, and maintaining the consistency of the external visible states of the master virtual machine and the slave virtual machines through a master-slave synchronization protocol. Compared with the prior art, according to the invention, an efficient master-slave synchronization protocol and an operating system cooperating with a virtual machine monitor are used; the method greatly reduces the tail time delay of user request processing, reduces the frequency of state difference caused by uncertainty of redundantoperation of the master and slave virtual machines, reduces the expenditure of state synchronization of the master and slave virtual machines, remarkably improves the overall throughput of an application program in a fault-tolerant system, and remarkably reduces the delay of request processing.

Description

technical field [0001] The present invention relates to the technical field of virtual machines, and in particular, to a method and system for realizing fault tolerance of virtual machines. Background technique [0002] Virtual machine technology is widely used in today's various cloud services. By virtualizing fixed and limited physical computer resources into configurable, controllable and easy-to-manage virtual computer resources, it makes user purchase policies more flexible and flexible, and improves the utilization and efficiency of physical computing and storage resources. , so that the server management and operation and maintenance costs are greatly reduced. [0003] Due to the advantages of convenient management and cost-effectiveness of virtual machine cloud products, as its development continues to mature, more and more applications, Internet services, database systems, etc. are deployed in virtual machines. Among them are mission-critical applications that req...

AI Technical Summary

Problems solved by technology

The fault-tolerant logic of this type of system is relatively simple, and the impact on the performance of the virtual machine is also low, but the fault recovery requires a long time to start the virtual machine, and at the same time, the network stack connection status, running memory status and other information in the virtual machine are all lost. , the services provided by the virtual machine will be affected

The second type of virtual machine fault-tolerant system improves the high availability of virtual machine services by backing up all the runtime information of the virtual machine in real time. However, in order to ensure the external consistency (External Consistency) of the virtual machine, all the It can only be sent to the client after the whole machine is backed up, which greatly increases the response delay

This can still result in significant performance penalties for applications that frequently produce indeterminate results when executing (for example, transaction processing differences due to the order in which threads are scheduled to execute and the order in which locks are acquired)

Images