M:N Failover Resource Management for Scalable High Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high availability systems rely on expensive and difficult-to-scale 1:1 physical redundant failover configurations, requiring extensive engineering efforts and formal hardware/software updates, especially when hardware components reach end of life, leading to challenges in maintaining system reliability and scalability.
Innovation Solution
A system dynamically load-balances redistribution elements across a group of computing resources using an M:N working configuration, accessing operational data from a central data store to identify load-balancing opportunities, select redistribution targets, and redeploy elements to maintain high availability without the need for 1:1 physical redundancy, utilizing a pool of multiple computing resources to achieve and maintain required availability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 1:1 physical redundant failover configuration is used, then system reliability is maintained, but hardware cost and device complexity increase significantly
Solution Approach 1:
The patent uses virtualization to create a virtual copy of the standby controller that can serve multiple primary controllers simultaneously. Instead of requiring physical 1:1 redundant hardware, a single physical standby controller is virtualized to provide backup services to N primary controllers, reducing hardware complexity while maintaining reliability through virtual redundancy
Solution Approach 2:
The standby controller is designed with multi-functionality to serve multiple primary controllers across different process areas. A single standby controller can assume the role of any failed primary controller through dynamic configuration, making the redundancy system universal rather than dedicated to specific hardware pairs
2Reliability
If 1:1 physical redundant failover configuration is used, then high availability is achieved, but scalability and adaptability deteriorate
Solution Approach 1:
The patent implements dynamic failover configuration where the standby controller can adaptively assume different roles based on which primary controller fails. The system dynamically reconfigures process area assignments and communication routes, allowing the same hardware to provide high availability across evolving system configurations without requiring physical reconfiguration
Solution Approach 2:
The patent introduces a virtualization dimension that separates logical redundancy relationships from physical hardware relationships. By operating in the virtualization layer, the system achieves scalable high availability where additional primary controllers can be added without proportionally increasing physical standby hardware, as virtual standby instances are created through software
3Reliability
If extensive engineering efforts and formal hardware updates are performed, then system reliability is maintained, but loss of time and productivity decrease
Solution Approach 1:
The system implements automated failover detection and execution mechanisms that operate without extensive manual engineering intervention. The standby controller automatically detects primary controller failures, assumes control, and restores system operation, eliminating the need for manual hardware updates and reducing maintenance time while preserving reliability
Data Source
AI summary
A system for dynamically load-balancing at least one redistribution element across a group of computing resources that facilitates at least an aspect of an Industrial Execution Process in an M:N working configuration is illustrated. Such a system may be configured to: access from a central data store, a configuration component operational data and capabilities or characteristics associated with the M:N working configuration; identify a load-balancing opportunity to trigger redistribution of a redistribution element to a redistribution target selected from a redistribution target pool defined by remaining computing resource components associated with the M:N computing resource working configuration; select at least one redistribution target for redeployment; redeploy the at least one redistribution element to the redistribution target; determine redeployment to the at least one selected redistribution target to be a viable redeployment; and execute the Industrial Execution Process utilizing the at least one redistribution element at the selected redistribution target.