SDN Load Balancing with Local Redistribution and Fewer Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing load balancing methods in Software-Defined Networks (SDNs) struggle to efficiently adapt to real-time traffic conditions and address load imbalance issues without requiring extensive global reconfiguration, leading to inefficient network utilization.
Innovation Solution
A network device with local load balancing capabilities that detects and adapts to load imbalances by modifying forwarding rules to redistribute flows among output ports, using measured or predicted loads to minimize deviations from target loads, and employing techniques like Variable Sized Bin Packing Problem (VSBPP) to optimize load distribution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If centralized control plane entities continuously optimize routing and load balancing using global network view, then network utilization is improved, but device complexity and computational demand increase significantly
Solution Approach 1:
The patent segments the load balancing function into two parts: (1) centralized controller that provides initial flow distribution and global optimization, and (2) distributed network devices that perform local load monitoring and adaptive flow redistribution. This segmentation reduces the computational burden on the centralized controller while maintaining overall network optimization.
Solution Approach 2:
The patent enables network devices to perform local load balancing decisions based on real-time traffic conditions observed at each device. Each device independently monitors its own output port loads and redistributes flows locally without requiring continuous centralized intervention, making the system adaptable to local conditions while reducing controller complexity.
2Manufacturing precision
If extensive global reconfiguration is performed to address load imbalance issues, then load balancing accuracy is improved, but adaptation speed and response time deteriorate
Solution Approach 1:
The centralized controller pre-configures flow distribution rules based on global network state before traffic patterns become problematic. Network devices then monitor traffic conditions and only perform local redistributions when imbalances are detected, avoiding the need for continuous global reconfiguration and enabling faster response to changing conditions.
Solution Approach 2:
The system transitions from static centralized configuration to dynamic local adaptation. Network devices continuously monitor traffic conditions and autonomously redistribute flows in response to real-time load changes, enabling the system to adapt quickly to dynamic traffic patterns without waiting for centralized reconfiguration.
3Manufacturing precision
If individual forwarding rules are created for each flow to achieve precise load balancing, then load distribution precision is improved, but the number of forwarding rules and memory usage increase
Solution Approach 1:
The patent merges multiple flows into flow aggregates that share common characteristics (source, destination, protocol). Instead of creating individual forwarding rules for each micro-flow, the system creates rules at the aggregate level while still achieving fine-grained load balancing by selectively redistributing specific aggregates based on their contribution to load imbalances.
Solution Approach 2:
The system applies load balancing actions selectively to only those flow aggregates that contribute significantly to load imbalances, rather than applying rules to all flows. This partial action approach achieves effective load balancing while minimizing the number of forwarding rules required in the distributed system.
Data Source
Figure 1~2
Figure 3
Figure 4A
AI summary
The present disclosure relates to a device and method for a traffic forwarding network device and proposes a solution for imbalance issues by adapting load balancing to real traffic conditions. The network device tries to solve imbalance issues locally by readjusting the traffic of problematic flows and in case the issues cannot be solved locally, notifies a central network controller to reconfigure the network in order to solve the imbalance issue.