Dynamic Load Balancing in Network Aggregation Groups

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing hash-based approaches for data flow management in aggregation groups lead to data congestion and inefficient bandwidth utilization due to imbalances in data traffic across members, resulting in oversubscription and idle paths.

Innovation Solution

A system and method for dynamic load balancing that reassigns data flows based on loading conditions and inactivity thresholds, using quality metrics and quality/loading bands to select optimal paths within aggregation groups, and optionally utilizing alternative higher-cost paths when primary paths are overloaded.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If hash-based schemes are used to map data flows to aggregation group members, then data flows can be distributed across multiple paths, but traffic imbalance occurs leading to oversubscription and idle paths

Engineering Contradiction:
Improvebandwidth utilizationVSAvoidtraffic distribution balance
Core Design Contradiction:
ProductivityVSEase of operation

Solution Approach 1:

The patent implements dynamic load balancing by continuously monitoring traffic conditions and reassigning data flows between aggregation group members based on current loading states. This transforms the static hash-based assignment into a dynamic system that adapts to changing traffic patterns, preventing both oversubscription and idle paths

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system employs feedback mechanisms where traffic loading information from aggregation group members is continuously collected and used to make reassignment decisions. This closed-loop control enables the system to respond to traffic imbalances by redistributing flows from overloaded members to underutilized members

Inventive Principle:
Principle #23Feedback

2Ease of operation

If reassignment of data flows is performed dynamically, then traffic balance improves, but packet reordering may occur

Engineering Contradiction:
Improvetraffic distribution balanceVSAvoidpacket delivery order
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent applies preliminary action by waiting for data flows to become inactive or idle before reassigning them to different aggregation group members. This proactive approach ensures that reassignment only occurs when no packets are currently being transmitted, thereby preventing packet reordering while still enabling dynamic load balancing

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses preliminary anti-action by preventing reassignment during active transmission periods. By detecting when data flows are active and blocking reassignment during those periods, the system proactively avoids the harmful effect of packet reordering before it can occur

Inventive Principle:
Principle #9Preliminary anti-action

3Productivity

If only least-cost paths are used in aggregation groups, then routing efficiency is maximized, but alternative paths remain underutilized

Engineering Contradiction:
Improverouting efficiencyVSAvoidpath utilization flexibility
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent makes aggregation group members universal by enabling both least-cost and alternative paths to serve as active transmission members. Instead of restricting aggregation groups to only least-cost paths, the system allows alternative paths to function equally as valid transmission routes, improving overall path utilization while maintaining routing efficiency through intelligent flow assignment

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS10057174B2Dynamic group multipathing
Publication Date: 2018.08.21 AVAGO TECHNOLOGIES GENERAL IP (SINGAPORE) PTE LTD
  • US10057174B2 patent drawing
  • US10057174B2 patent drawing
  • US10057174B2 patent drawing

AI summary

The present disclosure is directed to a system and method for dynamic and adaptive load balancing. An example method includes receiving, at a network device, a data packet to be sent via one or more aggregation groups, where the aggregation groups each comprise a plurality of aggregate members. The example method further includes determining, based on the data packet, a flow identifier of a flow to which the data packet belongs and determining a state of the flow. The example method also includes determining, based on the flow identifier and the state of the flow, an assigned member of the plurality of aggregate members for the flow and communicating the packet via the assigned member.