Traffic Load Balancing via Dynamic Distribution Formulas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing load balancing methods for network traffic distribution across multiple processors fail to achieve even distribution, especially in dynamically changing traffic patterns, leading to inefficiencies in processing loads across IPSs and other network security systems.
Innovation Solution
A two-stage load balancing method that selects and applies distribution formulas to associate packets with output links, using a pointer system to monitor and adjust load distribution parameters, ensuring even processing loads across multiple processors by dynamically reassigning packets based on current traffic patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If hashing on source and destination address fields is used to distribute traffic load equally across multiple processors, then traffic distribution is simplified, but the resulting distributions vary at each processor based on traffic patterns leading to uneven load balancing
Solution Approach 1:
The system dynamically selects from multiple distribution formulas based on current traffic patterns. Instead of using a fixed hashing method, the load balancer monitors traffic characteristics and adapts the distribution strategy in real-time, allowing the system to maintain effectiveness as traffic patterns change while keeping the operational complexity manageable through automated formula selection
2Device complexity
If a fixed distribution method is used to associate packets with output links, then system complexity is reduced, but the system cannot adapt to dynamically changing traffic patterns
Solution Approach 1:
The load balancer is designed with multiple distribution formulas and a selection mechanism that can adapt to different traffic patterns. This multi-functional approach allows the same system to handle various types of traffic (HTTP, FTP, SMTP, etc.) and dynamic patterns without requiring separate specialized systems, achieving adaptability while managing complexity through a unified architecture
Solution Approach 2:
The system changes the distribution parameters by selecting different formulas based on traffic characteristics. When traffic patterns change, the system adjusts which distribution formula is applied, effectively changing the distribution parameters dynamically. This allows adaptation to varying traffic conditions without increasing fundamental system complexity
3Productivity
If multiple distribution formulas are monitored and selected based on traffic patterns, then load balancing effectiveness is improved, but system complexity increases
Solution Approach 1:
The load balancer automatically monitors traffic patterns and selects the appropriate distribution formula without external intervention. The system self-adjusts by evaluating current traffic characteristics and choosing the best formula from its set, maintaining high load balancing effectiveness while managing complexity through automation rather than manual configuration or complex control mechanisms
Data Source
AI summary
A method for load balancing selects one of a set of distribution formulas to associate packets of each of multiple communications sessions traversing a link to one of a plurality of output links and assigns the associated packets of at least one of the communications sessions to a different output link.