Network Flow Affinity Across Processing Planes for Lower Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network devices struggle with inefficient processing of network traffic flows due to the lack of effective techniques that leverage network flow affinity or locality, leading to suboptimal data communication speeds.
Innovation Solution
Implementing a network device with multiple processing planes, where the same processing component that initially processed a network flow is utilized for subsequent processing, leveraging flow affinity and locality to enhance processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple processing planes are used to handle network flows, then processing capacity and throughput are improved, but flow processing latency increases due to the need to switch between processing planes
Solution Approach 1:
The control plane performs preliminary actions by programming the data plane with flow handling instructions before actual data packets arrive. This includes installing flow entries in forwarding tables, configuring packet processing actions, and preparing processing components in advance so that when flows arrive, the data plane can execute pre-configured operations without delay.
Solution Approach 2:
The system segments network flow processing into distinct control plane and data plane functions. The control plane handles flow identification, decision-making, and programming of processing components, while the data plane executes packet forwarding and processing operations. This segmentation allows each plane to be optimized independently and enables parallel processing of multiple flows.
2Speed
If flow affinity is leveraged to reuse the same processing component for subsequent packets, then processing speed is improved, but device complexity increases due to tracking and managing flow affinity mappings
Solution Approach 1:
The data plane performs self-service by autonomously identifying flow affinity and redirecting packets to the appropriate processing component without requiring continuous control plane intervention. The system uses flow affinity mappings stored in the data plane to automatically route subsequent packets of the same flow to the same processing component, reducing the need for complex centralized management.
Solution Approach 2:
The system introduces flow affinity mappings as an intermediary mechanism between the control plane and data plane. These mappings store the relationship between flows and processing components, enabling the data plane to make intelligent routing decisions based on pre-established affinity relationships without requiring complex real-time coordination.
3Productivity
If cached data structures are used to store flow information, then processing efficiency is improved, but memory resource consumption increases
Solution Approach 1:
The system applies local quality by caching flow information at specific locations within the processing architecture where it is most needed. Flow affinity mappings and related metadata are stored in memory structures accessible to both control and data planes, enabling fast retrieval and processing. The caching strategy focuses on storing only essential flow attributes rather than complete flow state information.
Data Source
AI summary
Improved network traffic flow processing techniques are described. In a network device providing multiple processing planes, each processing plane comprising multiple processing units, techniques are described that take advantage of flow affinity/locality principles such that the same processing component of a processing plane, which previously performed processing for a network flow, is used for performing subsequent processing for the same network flow. This enables faster processing of network traffic flows by the network device. In certain implementations, the techniques described herein can be implemented in a network virtualization device (NVD) that is configured to perform network virtualization functions.