Fabric Data Transport Scheduling via Virtual Output Queues

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Modern data networks face inefficiencies and complexity in fabric-based architectures due to high speedup requirements and complex scheduling mechanisms, leading to the need for large buffers and complex fabric designs to ensure efficient data flow and Quality of Service (QoS) guarantees.

Innovation Solution

Implementing a system where input controllers schedule data packet transport across a fabric based on permission information received from output controllers, using transmission credits that consider data flow status, available storage, and QoS profiles, allowing for pre-scheduling of data packets and optimizing data flow without requiring complex fabric operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a fully output buffered architecture is employed to ensure work conserving operation and simplify packet scheduling, then reliability and ease of operation are improved, but device complexity and cost increase significantly for mid to high capacity switches

Engineering Contradiction:

Improvework conserving operationVSAvoidbuffering architecture complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the buffering function by introducing virtual output queues (VOQs) at input ports, separating packet waiting from physical buffer location. This allows input ports to buffer packets virtually before fabric transmission, eliminating the need for large physical buffers at output ports while maintaining work conserving operation through virtual queue management.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a fabric controller as an intermediary component that manages packet transmission across the fabric. This controller coordinates between input and output ports, regulating data flow and preventing congestion without requiring large buffers at either end, thus simplifying the overall buffering architecture while maintaining reliability.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If high speedup is implemented in fabric-based architecture to support entire switching device capacity, then productivity is improved, but device complexity and buffer requirements increase

Engineering Contradiction:

Improvedata transport capacityVSAvoidfabric architecture complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent implements preliminary action by having input ports prepare packets in virtual output queues before actual fabric transmission. Packets are classified, prioritized, and organized in advance based on destination and QoS requirements, allowing the fabric to transmit data more efficiently without requiring high speedup ratios, thus reducing fabric architecture complexity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces dynamic packet scheduling based on real-time fabric status and QoS profiles. The system adapts transmission priorities and routing decisions dynamically rather than using static high speedup architecture, allowing the fabric to achieve high productivity through intelligent control rather than excessive hardware capacity, thereby reducing device complexity.

Inventive Principle:

Principle #15Dynamics

3Reliability

If complex scheduling mechanisms are used to ensure QoS guarantees in fabric-based networks, then reliability is improved, but device complexity and manufacturing difficulty increase

Engineering Contradiction:

ImproveQoS guaranteesVSAvoidscheduling mechanism implementation

Core Design Contradiction:

ReliabilityVSEase of manufacture

Solution Approach 1:

The patent applies local quality by implementing QoS scheduling independently at each input port through virtual output queues. Each input port manages its own packet prioritization and queuing based on local traffic characteristics and QoS profiles, rather than requiring complex centralized scheduling. This distributed local management simplifies implementation while ensuring reliable QoS guarantees through consistent per-port policy enforcement.

Inventive Principle:

Principle #3Local quality

4Reliability

If large buffers are deployed at output controllers to adapt traffic rate, then reliability is improved, but device complexity and buffer requirements increase

Engineering Contradiction:

Improvetraffic rate adaptationVSAvoidbuffer capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts the buffering function from output controllers and relocates it to input ports through virtual output queues. By taking out the buffer requirement from the output side and placing it at the input side, the system achieves traffic rate adaptation and congestion prevention without requiring large buffers at output controllers, thus reducing overall buffer capacity requirements while maintaining reliability.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS7619970B2Method, device and system of scheduling data transport over a fabric

Publication Date: 2009.11.17 AVAGO TECHNOLOGIES INTERNATIONAL SALES PTE LTD

AI summary

Embodiments of the invention provide systems, devices and methods to schedule data transport across a fabric, e.g., prior to actual transmission of the data across the fabric. In some demonstrative embodiments, a packet switch may include an input controller to schedule transport of at least one data packet to an output controller over a fabric based on permission information received from the output controller. Other embodiments are described and claimed.