Load Balancing Virtual Edge Systems for High Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In high availability computing systems, the failover system often remains idle and consumes resources unnecessarily, wasting memory and processing power, as it waits for potential takeover from the primary system, even when implemented as a virtual element.
Innovation Solution
Implementing a method for load balancing between a pair of virtual edge systems by assigning a virtual network address, generating and sharing state information, and directing network traffic between them, allowing both systems to process network traffic and act as both active and failover nodes, thereby utilizing resources more efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the failover system is kept powered on and operational to ensure immediate takeover capability, then high availability and service continuity are improved, but computing resources (memory and processing power) are wasted unnecessarily
Solution Approach 1:
The patent implements dynamic role assignment where edge systems can switch between active and standby roles based on operational needs. The load balancer dynamically directs traffic to either the active edge system or the standby edge system, allowing resources to be optimized while maintaining failover capability. This resolves the contradiction by making the system flexible rather than statically keeping the standby system fully operational.
Solution Approach 2:
The standby edge system performs self-updates and maintenance tasks during its standby period without requiring full operational resources. The system can apply patches, update configurations, and prepare for potential failover while consuming minimal resources, thus reducing waste while maintaining reliability.
2Loss of energy
If the standby edge system is kept dormant to save resources, then resource utilization is improved, but takeover time increases when failover is needed
Solution Approach 1:
The standby edge system performs preliminary actions during its standby state, including pre-loading configurations, validating system integrity, and preparing network interfaces. This ensures that when failover is triggered, the system can transition rapidly without full initialization delays, thus reducing takeover time while maintaining resource efficiency.
Solution Approach 2:
The system implements continuous health monitoring and heartbeat mechanisms between the active and standby edge systems. The load balancer receives feedback about system status and can proactively trigger failover or initiate resource allocation before complete failure occurs, reducing effective downtime while optimizing resource usage during normal operation.
3Device complexity
If a single active edge system handles all traffic to simplify management, then system complexity is reduced, but resource utilization efficiency decreases
Solution Approach 1:
The load balancer serves multiple functions: traffic distribution, health monitoring, failover management, and resource optimization. By centralizing these functions in the load balancer, the system maintains simplicity in management while enabling efficient resource utilization across multiple edge systems. The load balancer can dynamically adjust traffic distribution based on system capacity and demand.
Solution Approach 2:
The patent segments traffic handling responsibilities between the load balancer and multiple edge systems. The load balancer handles traffic distribution and control plane functions, while edge systems handle data plane processing. This segmentation allows multiple edge systems to be utilized efficiently without significantly increasing management complexity, as the load balancer abstracts the complexity of coordinating multiple systems.
Data Source
AI summary
The technology disclosed herein enables load balancing between a pair of virtual edge systems configured for high availability at an edge of a local network environment. In a particular embodiment, a method provides assigning a virtual network address to the pair of virtual edge systems. The method further provides generating state information used by one or more stateful functions of a first virtual edge system of the pair of virtual edge systems and transferring the state information to a second virtual edge system of the pair of virtual edge systems. Also, the method provides directing a first portion of network traffic to the first virtual edge system and a second portion of the network traffic to the second virtual edge system. The network traffic comprises packets addressed with the virtual network address.