Virtual Instance Failover via Hosted Service System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high initial costs and complexity of deploying enterprise application software with traditional three-tiered architecture pose a barrier for small and medium businesses, while existing cloud computing solutions lack effective failover and recovery mechanisms for computing applications hosted on virtual instances, compromising reliability and security.
Innovation Solution
A system and method for failover and recovery of computing applications hosted on virtual machines, utilizing a hosted service system that provides automated administration, encrypted networking, and secure integration between private and public virtualization environments, enabling detection of failed instances and automatic restoration to a last known operational state.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional three-tiered architecture is deployed locally, then reliability and security are maintained, but deployment costs and complexity increase substantially
Solution Approach 1:
The patent introduces a hosted service system as an intermediary between the client and the virtualization service provider. This hosted service system manages the computing application and handles failover operations, abstracting away the complexity of direct local deployment while maintaining reliability through centralized management and automated recovery mechanisms.
Solution Approach 2:
The patent creates virtual instances as copies of the computing application that can run on remote virtualization services. These virtual instances replicate the application's functionality and state, allowing failover to occur by switching between instances without requiring complex local infrastructure, thus reducing deployment complexity while maintaining reliability.
2Productivity
If cloud computing is used to reduce costs, then scalability and maintainability improve, but failover and recovery mechanisms are insufficient
Solution Approach 1:
The patent implements preliminary failover actions by detecting failures early through monitoring mechanisms and automatically initiating recovery procedures. The system performs preliminary assessments of instance health and pre-configures failover targets, enabling rapid recovery without compromising scalability or requiring complex manual intervention.
Solution Approach 2:
The patent establishes feedback loops through instance status monitoring and health checks that continuously report on the state of virtual instances. This feedback mechanism enables the hosted service system to detect failures, trigger appropriate failover actions, and verify recovery success, thereby enhancing reliability while maintaining the scalability benefits of cloud computing.
3Reliability
If automated failover is implemented, then reliability improves, but system complexity increases
Solution Approach 1:
The patent implements self-service automation where the hosted service system autonomously monitors instance health, detects failures, and executes failover operations without requiring manual intervention. The system automatically manages the complexity of failover logic, health checking, and instance restoration, thereby improving availability while preventing the complexity from propagating to the user level.
Data Source
AI summary
A system and method are described for performing failover and recovery of a computing application hosted by a virtual instance of a machine. An example system may include an instance status monitor, an instance failure detector, and a failover module. The instance status monitor may be configured to monitor status of a virtual instance hosted by a virtualization service. The instance failure detector may be configured to detect that the virtual instance has failed or is running in a degraded or failed state and is thus to be considered a failed virtual instance. The failover module may be configured to facilitate launching a new virtual instance in a recovered state that reflects a last known operational state of the failed virtual instance.