Predictive Network Routing Controller for Traffic Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hardware load balancers struggle with efficiently distributing network traffic across geographically disparate servers, often resulting in inefficient routing and single points of failure.
Innovation Solution
The integration of load balancing and routing decisions, where both the route and destination are determined together by a controller, allowing for holistic decision-making based on factors like network congestion, server utilization, and geographic diversity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional hardware load balancers are used to distribute network traffic, then traffic distribution capability is provided, but routing efficiency deteriorates due to hair-pinning traffic back outward into the wide area network
Solution Approach 1:
The patent combines load balancing functionality with routing functionality into a single integrated system. The load balancer determines both the destination server and the optimal network route simultaneously, eliminating the need for separate routing decisions that cause hair-pinning inefficiencies. This merging of functions allows traffic to be distributed to geographically disparate servers through direct routes rather than back-and-forth paths.
2Device complexity
If a single instance of hardware load balancer is used, then device simplicity is maintained, but system reliability deteriorates due to single points of failure
Solution Approach 1:
The patent segments the load balancing function into multiple geographically distributed instances. Each load balancer instance serves a specific region or set of servers, and the system as a whole provides failover capabilities. If one load balancer instance fails, other instances continue to handle traffic for their respective regions, eliminating single points of failure while maintaining manageable complexity through modular deployment.
3Adaptability or versatility
If traffic is routed to geographically disparate servers, then service coverage is improved, but routing efficiency deteriorates without integrated route determination
Solution Approach 1:
The patent performs route determination in advance as part of the load balancing decision process. When a load balancer determines the optimal destination server based on geographic proximity and server availability, it simultaneously determines the optimal network route to that destination. This preliminary action ensures that traffic is routed efficiently from the start, avoiding the need for subsequent route changes or back-paths that would reduce efficiency.
Data Source
AI summary
Methods and apparatuses are described for routing of network traffic based on future utilization of a plurality of network devices. For example, future utilization for each of a plurality of network devices may be predicted. Then, based on the predicted future utilization, a route may be determined that comprises at least some of the plurality of network devices. An instruction to route data packets based on the route can then be sent to one (e.g. the first) of the network devices in the determined route.