Controller-Based Port Path Determination for Data Center Networks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Large-scale data center networks face limitations in capacity due to the size of MAC and routing tables, which restrict network expansion and efficiency.

Innovation Solution

A method and apparatus that utilize a controller to obtain port path information by determining absolute port paths through network topology, allowing packet forwarding without relying on MAC or routing entries, thereby enhancing network capacity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If MAC table or routing table is used for packet forwarding, then packet forwarding function is achieved, but network capacity is limited due to table size constraints

Engineering Contradiction:
Improvenetwork capacityVSAvoidMAC table size
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts the forwarding decision logic from the traditional MAC table/routing table approach. Instead of storing complete MAC addresses and routing information in large tables, the system extracts only essential port path information (inbound port, outbound port, and path identifier) that is needed for forwarding decisions. This dramatically reduces the table size while maintaining forwarding functionality.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the forwarding information into multiple components: path identifier, inbound port, outbound port, and next-hop information. Each component is stored separately and used in combination for forwarding decisions. This segmentation allows the system to avoid storing complete MAC address tables while still achieving accurate packet forwarding through piecewise information reconstruction.

Inventive Principle:
Principle #1Segmentation

2Productivity

If MAC table or routing table size is increased to support larger networks, then network capacity increases, but device resource consumption increases

Engineering Contradiction:
Improvenetwork capacityVSAvoiddevice resources
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

The patent employs a lightweight forwarding information structure that consumes minimal device resources. Instead of maintaining large, resource-intensive MAC tables, the system uses compact path-based forwarding entries that require significantly less memory and processing power. The port path information structure is designed to be resource-efficient while supporting large-scale networks.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

3Loss of information

If port path information is obtained through traditional routing protocols, then routing information is acquired, but information overhead and processing complexity increase

Engineering Contradiction:
Improverouting information completenessVSAvoidprocessing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent extracts only the essential port path information needed for forwarding from the complete routing information. Instead of processing and maintaining full routing tables with complete path details, the system extracts and stores only the critical elements (path identifier, inbound/outbound ports) required for efficient forwarding decisions, reducing both information overhead and processing complexity.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3253012B1Method and apparatus for obtaining port path
Publication Date: 2020.05.20 HUAWEI TECH CO LTD
  • EP3253012B1 patent drawingFigure 1
  • EP3253012B1 patent drawingFigure 2
  • EP3253012B1 patent drawingFigure 3

AI summary

Embodiments of the present invention provide a method for obtaining a port path and an apparatus, to help improve a network capacity. The method includes: receiving, by a controller, a request message from a first server, where the request message is used to request port path information, and the port path information includes a port that a logical link from the first server to a second server passes through; obtaining, by the controller, a first APP and a second APP according to network topology information, an identifier of the first server, and an identifier of the second server, where the first APP includes a port that a logical link from a root node to the first server passes through, and the second APP includes a port that a logical link from the root node to the second server passes through; and obtaining, by the controller, the port path information according to the first APP and the second APP, and sending the port path information to the first server.