Dynamic Traffic Orchestration Switch System

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Traditional networking systems lack dynamic orchestration capabilities to efficiently manage and direct data packets within data centers, leading to suboptimal traffic distribution and server utilization.

Innovation Solution

The implementation of a switch system with a master and slave configuration, where the master switch actively processes packets and directs them to servers, and the slave switch stands by, modifying packet destinations and headers to ensure efficient traffic routing and return paths, utilizing switching engines and server logic to manage policy headers and traffic flow.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional routing methods using MAC addresses and VLAN identifiers are used, then packet forwarding is simple and direct, but dynamic traffic orchestration and server utilization optimization are lacking

Engineering Contradiction:
Improvedynamic traffic orchestration capabilityVSAvoidswitching system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

A policy server is introduced as an intermediary component that manages traffic orchestration policies and communicates with switching engines. The policy server stores policy information in a database and provides dynamic routing decisions without requiring complex logic in the switches themselves, thus enabling adaptability while controlling device complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements dynamic traffic orchestration by allowing routing policies to be modified in real-time based on server utilization, traffic patterns, and operational conditions. The switching engines receive updated policy information from the policy server, enabling the system to adapt to changing conditions without physical reconfiguration.

Inventive Principle:
Principle #15Dynamics

2Reliability

If a single active switch is used for packet processing, then system operation is simplified, but system reliability and fault tolerance are reduced

Engineering Contradiction:
Improvesystem fault toleranceVSAvoidswitch configuration complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system is pre-configured with multiple switches in a virtual chassis where one switch is designated as active and others as standby before any failure occurs. The standby switches are pre-loaded with policy information and ready to take over immediately if the active switch fails, ensuring continuous operation without requiring complex runtime decision-making.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements monitoring of the active switch's operational status with automatic feedback mechanisms. When the active switch fails, the standby switches detect the failure through feedback signals and automatically promote one of the standbys to active status, maintaining system reliability through automated failover procedures.

Inventive Principle:
Principle #23Feedback

3Productivity

If packets are routed through multiple switches for processing, then traffic distribution is improved, but packet processing time and latency increase

Engineering Contradiction:
Improvetraffic distribution efficiencyVSAvoidpacket processing time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system segments traffic handling by function: the first switch handles ingress and initial policy application, intermediate switches handle specific processing tasks, and the last switch handles egress. This segmentation allows parallel processing of different packet components simultaneously, improving traffic distribution efficiency while minimizing sequential processing delays.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Policy information and routing decisions are prepared and cached in advance by the policy server and switching engines before packets arrive. This preliminary action allows switches to make rapid forwarding decisions without real-time computation delays, reducing packet processing time while maintaining efficient traffic distribution through pre-computed optimal paths.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS9426067B2Integrated switch for dynamic orchestration of traffic
Publication Date: 2016.08.23 KYNDRYL INC
  • US9426067B2 patent drawing
  • US9426067B2 patent drawing
  • US9426067B2 patent drawing

AI summary

A network element within a data center comprises a switch adapted to receive a data packet, a server in communication with the switch, and a switching engine in the switch. The switching engine may be configured to modify policy header information in the data packet to direct the packet to the server. The server may be configured to modify the policy header information while the data packet is in the server to return the data packet to said switch.