Network Teaming Split Across Switches for Redundant Load-Balancing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional port-trunking techniques limit the ability to split network resources of an SLB team between different switches, forcing users to choose between network redundancy and receive traffic aggregation and load-balancing, thereby restricting the benefits of redundant connections.

Innovation Solution

The system allows for the configuration of network resources into teams that can be split across multiple switches, enabling redundant connections and receive traffic aggregation and load-balancing by designating one group as an SLB sub-team within a TLB team, with the team MAC address being dynamically reassigned to maintain connectivity and throughput during failovers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If network resources are split between different switches for redundancy, then network reliability is improved, but receive traffic aggregation and load-balancing capability deteriorates

Engineering Contradiction:
Improvenetwork redundancyVSAvoidreceive throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the network team into sub-teams, where each sub-team is associated with a specific switch. This allows the team to be distributed across multiple switches while maintaining receive load-balancing capabilities within each sub-team. The segmentation enables redundancy through multiple switches while preserving traffic aggregation through the sub-team structure.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements a nested structure where SLB sub-teams are nested within a parent TLB team. Each SLB sub-team handles receive load-balancing for its associated switch, while the parent TLB team provides overall transmit load-balancing and team functionality. This nesting allows receive traffic aggregation across multiple switches while maintaining network redundancy.

Inventive Principle:
Principle #7Nested doll (Nesting)

2Stability of the object's composition

If conventional port-trunking techniques are used to maintain team integrity, then team stability is improved, but flexibility in resource allocation across switches deteriorates

Engineering Contradiction:
Improveteam integrityVSAvoidresource allocation flexibility
Core Design Contradiction:
Stability of the object's compositionVSAdaptability or versatility

Solution Approach 1:

The team is segmented into multiple SLB sub-teams, each with its own stability mechanisms through switch-specific associations. This segmentation allows flexible resource allocation across different switches while each sub-team maintains its own integrity. The parent TLB team provides overarching team functionality while allowing sub-team independence.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic MAC address assignment where the team MAC address can be dynamically reassigned between different sub-teams based on failover conditions. This dynamic behavior allows the team to adapt to switch failures or resource changes while maintaining overall team stability and continuity of service.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS7646708B2Network resource teaming combining receive load-balancing with redundant network connections
Publication Date: 2010.01.12 HEWLETT PACKARD ENTERPRISE DEV LP
  • US7646708B2 patent drawing
  • US7646708B2 patent drawing
  • US7646708B2 patent drawing

AI summary

A computer system teams its network resources by apportioning each of the one or more network resources into one of two or more groups. The resources of each of the groups have links that are operative to be coupled in parallel to a different one of two or more network devices that provide connectivity to a network. One of the two or more groups is selected to be a primary group. A team layer2 address is assigned to one of the resources apportioned to the selected group. Each of the resources of the selected group is enabled to receive on the team layer2 address. The network device to which the selected group is coupled is enabled to load balance data received from the network among the resources of the selected group. Data transmitted from the system to the network is load-balanced over all of the resources independent of the group to which they belong.