Virtual Machine File Failover for Cluster Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disaster recovery techniques require a significant number of physical machines at the disaster recovery site to be operational and powered on at all times, leading to high costs and increased resource consumption, particularly in terms of network bandwidth and memory, and necessitate physical to virtual conversion of machines, which is resource-intensive.
Innovation Solution
A method and apparatus that utilize preconfigured virtual machine files on a backup computer to ensure high availability of applications in a computer cluster without the need for physical to virtual conversion, allowing for seamless failover and failback of software applications between clusters using a global service group and data replicator, reducing the need for constant resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If physical machines are kept operational at the disaster recovery site, then application availability is improved, but resource consumption and costs increase
Solution Approach 1:
The patent uses virtual machine files (copies of the computing environment) instead of physical machines at the disaster recovery site. These virtual machine files can be quickly instantiated when needed, providing application availability without requiring continuous operation of physical infrastructure at the backup site.
Solution Approach 2:
The system dynamically activates virtual machines from stored files only when failover is required, rather than maintaining continuous operational state. This dynamic approach allows the disaster recovery infrastructure to transition from an active state (consuming resources) to a standby state (minimal resource consumption) while maintaining the capability to provide high availability.
2Adaptability or versatility
If physical to virtual conversion is performed, then device flexibility is improved, but processing time and network bandwidth are consumed
Solution Approach 1:
The patent performs the virtualization conversion in advance, creating virtual machine files from physical machines before any failover event occurs. These pre-converted virtual machine files are stored and can be quickly activated when needed, eliminating the need to perform time-consuming conversion during actual disaster recovery operations.
3Reliability
If virtual machines are kept ready for disaster recovery, then application availability is improved, but memory and network bandwidth usage increase
Solution Approach 1:
Instead of maintaining expensive, continuously running virtual machines with full resource allocations, the system uses inexpensive virtual machine files that can be quickly instantiated when needed. The virtual machines are created on-demand rather than maintained in a continuous ready state, reducing ongoing memory and network bandwidth consumption while still providing disaster recovery capability.
Data Source
AI summary
A method and apparatus for achieving high availability in a computer cluster is provided. In one embodiment, a method for failing over a software application to a virtual machine from a physical computing environment comprises processing at least one virtual machine file at a backup computer, wherein the at least one virtual machine file is preconfigured with a software application that is hosted by a computer cluster and in response to an occurrence a fault at the computer cluster, accessing the at least one virtual machine file to operate the software application.