Network Traffic Prioritization via Dynamic Load Balancing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing receive side scaling (RSS) technologies in computing platforms face inefficiencies in directing packets across multiple processors, leading to uneven processing loads and potential bottlenecks, as they rely on fixed hash functions and redirection tables that do not dynamically adapt to changing processor loads.

Innovation Solution

Implementing a DPC retargeting engine within the device driver that dynamically prioritizes and redistributes DPC management objects across multiple processors based on current processing loads, using a protocol prioritizing application to ensure higher priority traffic is processed efficiently and balancing the load among processors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If fixed hash functions and redirection tables are used to direct packets to processors, then packet distribution is simplified, but processing load becomes uneven and processors may become overloaded

Engineering Contradiction:

Improvepacket distribution mechanismVSAvoidnetwork processing throughput

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent implements dynamic load balancing by continuously monitoring processor load metrics and adjusting packet routing decisions in real-time. The system transitions from static hash-based routing to dynamic routing that adapts to changing processor states, ensuring optimal load distribution across processors while maintaining high network processing throughput

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent employs feedback mechanisms where the system monitors processor load metrics and uses this information to adjust packet routing decisions. The load balancer receives feedback about processor states and dynamically modifies routing behavior to prevent overload and maintain balanced processing across all processors

Inventive Principle:

Principle #23Feedback

2Productivity

If packets are distributed across multiple processors using fixed routing, then single processor bottleneck is avoided, but load imbalance occurs across processors

Engineering Contradiction:

Improvenetwork processing throughputVSAvoidload distribution balance

Core Design Contradiction:

ProductivityVSEase of operation

Solution Approach 1:

The system implements dynamic load balancing that continuously adapts to changing processor conditions. By monitoring load metrics and adjusting routing decisions in real-time, the system maintains balanced load distribution across processors while maximizing overall network processing throughput

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent changes routing parameters dynamically based on processor load conditions. Instead of using fixed hash functions, the system adjusts routing parameters such as destination processor selection based on current load metrics, ensuring optimal load distribution while maintaining high throughput

Inventive Principle:

Principle #35Parameter changes

3Device complexity

If all network traffic is processed equally across processors, then simple routing is maintained, but high-priority traffic cannot be expedited

Engineering Contradiction:

Improverouting logicVSAvoidquality of service for priority traffic

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent applies different processing qualities to different traffic types by implementing protocol-aware prioritization. High-priority protocols such as Fibre Channel are routed to processors with available capacity and given preferential processing, while lower-priority traffic is routed based on general load conditions, ensuring quality of service for critical protocols

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS7764709B2Prioritization of network traffic

Publication Date: 2010.07.27 MICRON TECHNOLOGY INC

AI summary

A plurality of packets associated with a plurality of protocols are received, wherein the plurality of packets are to be processed by a plurality of processors. Packets associated with a first protocol are sent to be processed by at least one of the plurality of processors before sending packets associated with a second protocol to be processed, in response to determining that the packets associated with the first protocol have a higher priority for processing than the packets associated with the first protocol.