System-Level High Availability Service for Transparent Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Achieving high availability and reliability in computer systems is costly and time-consuming due to the need for custom applications, operating systems, or hardware, which can delay deployment and are often proprietary.
Innovation Solution
A system-level high availability service that provides transparent and automatic fault detection and recovery without modifying applications or operating systems, using a layered approach with Availability Managers and Duration Interfaces to manage TCP connections and kernel state, enabling seamless failover between primary and backup servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If custom applications, operating systems, or hardware are used to provide high availability, then system reliability is improved, but development time and cost increase significantly
Solution Approach 1:
The patent segments the high availability functionality into separate system-level services (Availability Managers, Duration Interfaces, TCPF, KState) that operate independently from applications and the OS kernel. This allows high availability to be provided as a modular service layer, eliminating the need for custom application or OS development while maintaining reliability.
Solution Approach 2:
The patent introduces an intermediary service layer between applications and the OS kernel that handles fault detection, monitoring, and failover operations. This mediator layer (comprising Duration IF, Duration AM, TCPF, and KState) transparently manages high availability without requiring modifications to applications or the kernel, thus reducing development time while preserving reliability.
2Reliability
If custom applications, operating systems, or hardware are used to provide high availability, then system reliability is improved, but cost increases substantially
Solution Approach 1:
The patent creates universal system-level services that can be applied to any application without customization. The Availability Managers and Duration Interfaces provide generic fault detection and failover capabilities that work across different applications and scenarios, eliminating the need for expensive custom development while maintaining high reliability.
Solution Approach 2:
The system implements self-service mechanisms where the Availability Managers automatically detect faults, initiate failover procedures, and manage backup servers without human intervention. This automation reduces the need for expensive custom management systems and specialized personnel, lowering overall costs while maintaining system reliability.
3Ease of manufacture
If transparent system-level services are used to provide high availability, then ease of implementation is improved, but system complexity increases
Solution Approach 1:
The patent uses copying mechanisms where backup servers maintain copies of application states and configurations. The system-level services manage these copies transparently, providing failover capability without requiring complex custom implementation. The copying approach simplifies the implementation while the modular architecture manages the inherent complexity.
Data Source
AI summary
A set of system-level high availability services for computer systems, including a service that functions in general terms like an extension of the operating system. By providing High Availability (HA) at the system-level, modifications to the applications or the operating system kernel are not required.