Proxy-Based Point-to-Node Traffic Routing in Middleware
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional web server farms face challenges in efficiently directing traffic to specific nodes for validation, testing, or maintenance without disrupting production services or increasing the risk of human error.
Innovation Solution
The system employs a proxy infrastructure partnered with load balancer infrastructure to override traffic routing for specific clients, allowing traffic from point-to-node clients to be directed to a designated server or virtual container while maintaining normal production routing for other clients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional load balancing is used to distribute traffic across multiple servers, then production efficiency is maintained, but the ability to direct specific client traffic to a designated node for validation or maintenance is lost
Solution Approach 1:
The patent applies local quality by implementing client-specific routing rules that allow different traffic routing behaviors for different clients. Point-to-node clients receive customized routing to designated application servers, while other clients continue to receive load-balanced traffic, thus maintaining local optimization without compromising global production efficiency
Solution Approach 2:
The traffic routing system is segmented into different routing paths: one path for point-to-node clients that directs traffic to specific designated servers, and another path for regular clients that uses traditional load balancing. This segmentation allows simultaneous execution of validation/maintenance operations and production workloads without interference
2Reliability
If application servers are taken offline for validation or maintenance, then thorough testing can be performed, but service disruption occurs
Solution Approach 1:
The system introduces an intermediary routing mechanism that separates validation traffic from production traffic. A designated application server can simultaneously handle validation requests from point-to-node clients and production requests from other clients, eliminating the need to take servers offline for validation while maintaining service availability
Solution Approach 2:
The system performs preliminary routing configuration to designate specific application servers for point-to-node clients before validation begins. This preliminary setup allows validation operations to proceed on dedicated servers without disrupting production services, as the routing rules are established in advance
3Adaptability or versatility
If manual traffic routing configuration is used to direct traffic to specific nodes, then validation can be performed, but the risk of human error increases
Solution Approach 1:
The system implements self-service through automated routing rule generation and management. The load balancer automatically creates and manages routing rules based on client identification, eliminating manual configuration tasks and reducing the risk of human error while maintaining the capability to direct traffic to specific nodes for validation
4Productivity
If all traffic is load balanced across the server farm, then resource utilization is optimized, but specific validation or maintenance operations cannot be performed on individual nodes
Solution Approach 1:
The routing system is made dynamic by implementing conditions-based routing rules that can adapt to different client types and operational states. The load balancer dynamically determines whether to apply load balancing or point-to-node routing based on client identification and current operational requirements, allowing flexible resource allocation for both validation and production workloads
Data Source
AI summary
Systems and methods directing requests to a particular node in a multi-tiered middleware environment are provided. In one embodiment, traffic requests can be dynamically routed to a desired server, and not be load balanced, based on an identification of the client generating the traffic requests.


