Queue Management for Network Packet QoS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing QoS control mechanisms for Internet traffic, especially in network processors, face challenges with substantial processing overhead, difficulty in dynamically changing queue priorities, and limited scalability, particularly in software solutions without extensive hardware support.
Innovation Solution
A scalable and configurable queue management system implemented by a network processor using a queue manager to assert interrupts and a core processor to apply an interrupt mask, allowing dynamic priority adjustments and scalable queue numbers through memory-based management, reducing processing overhead and enabling flexible QoS processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If software-based QoS control is used, then flexibility in queue management is improved, but processing overhead increases substantially
Solution Approach 1:
The patent introduces a dedicated queue manager module as an intermediary component that handles all queue management operations. This module maintains separate state information for each queue (packet counts, priority levels) and provides standardized interfaces for QoS control, thereby reducing the processing burden on the core processor while maintaining software-based flexibility.
Solution Approach 2:
The queue management system is segmented into independent queue structures, each with its own state tracking and control mechanisms. This modular approach allows individual queues to be managed separately, reducing the complexity of overall queue management and enabling efficient software-based QoS control without requiring the core processor to handle all queue operations simultaneously.
2Adaptability or versatility
If dynamic queue priority changes are enabled, then QoS adaptability is improved, but system complexity increases
Solution Approach 1:
The queue manager implements dynamic priority adjustment capabilities where queue priorities can be modified at runtime based on network conditions and QoS requirements. The system maintains a priority structure that can be reconfigured without requiring system restart or complex reinitialization, enabling flexible QoS adaptation while managing complexity through dedicated management routines.
3Adaptability or versatility
If the number of queues is increased for better traffic classification, then QoS granularity is improved, but scalability is limited
Solution Approach 1:
The queue manager is designed with a universal structure that can accommodate any number of queues through dynamic data structures and array-based implementations. The same core management routines and interfaces work regardless of the number of queues, enabling the system to scale from a small number of queues for simple traffic classification to a large number of queues for fine-grained QoS control without requiring different system architectures.
Data Source
AI summary
Various embodiments are directed to scalable and configurable queue management for network packet traffic Quality of Service (QoS). In one or more embodiments, the queue management may be implemented by a network processor comprising a queue manager to assert interrupts indicating that one or more queues require service, and a core processor to apply an interrupt mask to a status register value identifying the one or more queues that require service and to provide service during a particular service cycle to only those queues that are not masked out. Other embodiments are described and claimed.