Packet Distribution in Hierarchical Managed Switching Elements

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Managed edge switching elements in networks often struggle to process packets, particularly unknown or multicast/broadcast packets, as they lack necessary entries in their forwarding tables, leading to inefficient packet forwarding across managed non-edge switching elements.

Innovation Solution

A method is introduced to distribute packet processing across managed non-edge switching elements using a hash function applied to packet headers, determining the appropriate switching element for further processing based on generated hash values and hash ranges.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If managed edge switching elements process all packets locally, then packet forwarding speed is improved, but device complexity increases and scalability deteriorates

Engineering Contradiction:
Improvepacket forwarding speedVSAvoidswitching element complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent segments packet processing responsibilities across different switching element types. Edge switching elements handle known unicast packets locally for fast forwarding, while non-edge switching elements handle unknown packets, multicast, and broadcast traffic. This segmentation allows each switching element to be optimized for its specific function, improving overall speed without requiring every element to handle all packet types independently.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces non-edge switching elements as intermediaries between edge switching elements and the network core. These intermediary elements perform complex processing tasks (unknown packet lookup, multicast/broadcast handling) that would otherwise burden edge switching elements, thereby reducing edge element complexity while maintaining fast local forwarding for simple packet types.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If managed edge switching elements handle unknown packets, then packet processing capability is improved, but memory requirements and device complexity increase

Engineering Contradiction:
Improvepacket processing capabilityVSAvoidforwarding table requirements
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the forwarding table storage responsibility. Edge switching elements only maintain forwarding tables for known unicast destinations, keeping their memory requirements low. Non-edge switching elements maintain comprehensive forwarding tables for unknown packets, multicast groups, and broadcast domains. This segmentation allows edge elements to be simple and scalable while still providing comprehensive packet handling capability through the distributed system.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Non-edge switching elements act as intermediaries that provide comprehensive forwarding table coverage for edge elements. When an edge element encounters an unknown packet, it forwards to a non-edge element that has the capability to handle such packets through its more complete forwarding information. This intermediary approach allows edge elements to maintain versatility without requiring them to store all possible forwarding information locally.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If packet processing is distributed across multiple switching elements, then scalability is improved, but packet forwarding overhead increases

Engineering Contradiction:
Improvenetwork scalabilityVSAvoidpacket forwarding overhead
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent applies local quality by optimizing packet handling at each switching element based on its type and location. Edge switching elements perform fast local forwarding for known packets without unnecessary hops. Non-edge switching elements perform specialized processing for specific packet types. This localized optimization ensures that packets are handled efficiently at the most appropriate element, minimizing overall forwarding overhead while enabling scalable distribution of processing responsibilities.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent implements preliminary action by having non-edge switching elements pre-compute and maintain forwarding information for unknown packets, multicast groups, and broadcast domains. When packets arrive at edge elements, the forwarding decisions for complex packet types have already been prepared by non-edge elements, reducing the real-time processing overhead during actual packet forwarding operations.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS8964528B2Method and apparatus for robust packet distribution among hierarchical managed switching elements
Publication Date: 2015.02.24 VMWARE INC
  • US8964528B2 patent drawing
  • US8964528B2 patent drawing
  • US8964528B2 patent drawing

AI summary

For a network that includes several managed edge switching elements and several managed non-edge switching elements that are for implementing a logical switching element, some embodiments provide a method of distributing packet processing across the several managed non-edge switching elements. The method receives a packet for processing through the logical switching element. Based on a determination that the packet needs to be processed by a managed non-edge switching element, the method determines a particular managed non-edge switching element of the several managed non-edge switching elements to forward the packet. The method forwards the packet to the particular managed non-edge switching element for the particular managed non-edge switching element to process the packet.