AI Training Data Flow Forwarding with SDN Traffic Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data center networks face traffic imbalance due to the use of traditional traffic balance mechanisms, leading to reduced network throughput and poor performance in AI cluster services, especially in scenarios with a small number of high-traffic data flows.
Innovation Solution
An SDN controller is employed to select forwarding paths based on network topology and task information, configuring flow tables to achieve load balance and traffic balance across the network, using a global perspective to optimize data flow forwarding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing traffic balance mechanism is used to forward data flows, then forwarding is simple, but traffic imbalance occurs on links reducing network throughput
Solution Approach 1:
An SDN controller is introduced as an intermediary between data flows and forwarding nodes. The controller globally optimizes path selection for AI training data flows based on network topology and flow characteristics, while forwarding nodes simply execute configured flow tables. This mediator resolves the contradiction by centralizing complex optimization logic, enabling traffic balance and high throughput without requiring complex local forwarding decisions.
Solution Approach 2:
The SDN controller performs preliminary path optimization before data flows are established. By pre-calculating optimal paths based on network topology and AI training communication patterns, the controller configures forwarding flow tables in advance. This preliminary action ensures traffic balance is achieved from the start, avoiding the need for complex real-time adjustments during flow transmission.
2Productivity
If traditional traffic balance mechanism is used, then implementation is straightforward, but AI cluster service performance deteriorates due to traffic imbalance
Solution Approach 1:
The SDN controller acts as a specialized intermediary that understands both network topology and AI training communication patterns. It optimizes paths specifically for AI data flows by identifying source and destination task nodes and selecting paths that balance traffic across the network. This targeted approach improves AI cluster service performance without requiring complex modifications to forwarding nodes themselves.
Solution Approach 2:
The system changes the parameter of path selection from static, simple routing to dynamic, topology-aware routing optimized for AI workloads. The SDN controller adjusts path parameters based on network conditions, flow characteristics, and topology information, enabling traffic balance that specifically addresses AI cluster performance requirements without overwhelming forwarding node complexity.
3Productivity
If SDN controller optimizes path selection based on topology and task information, then traffic balance is achieved, but control system complexity increases
Solution Approach 1:
The SDN controller is designed as a universal platform that handles multiple functions: topology management, flow information collection, path optimization, and flow table configuration. By consolidating these diverse functions into a single multi-functional controller, the system achieves traffic balance efficiency without proportionally increasing overall system complexity, as the controller replaces multiple specialized components.
Solution Approach 2:
The SDN controller automatically collects flow information from forwarding nodes and task nodes, then autonomously performs path optimization and flow table configuration without manual intervention. This self-service capability reduces the operational complexity burden, as the controller manages its own optimization processes using collected network and task information, achieving traffic balance efficiency through automated decision-making.
Data Source
Figure 1
Figure 2
Figure 3a
AI summary
Examples of the present disclosure provide a method and a device for forwarding a data flow, an SDN controller and a storage medium. The method is applied to the SDN controller, and includes: acquiring task information of an AI training task, wherein the task information comprises a communication model of the AI training task, an address of a source task node and an address of a destination task node, and the communication model is to indicate an AI training data transferring relationship between the source task node and the destination task node; selecting a forwarding path between the source task node and the destination task node based on topology information of a network and the communication model, wherein the topology information includes a topology structure, a link state and a utilization rate of link bandwidth; configuring a forwarding flow table to each forwarding node on the forwarding path, causing the each forwarding node to forward a data flow of the AI training task from the source task node to the destination task node along the forwarding path based on the forwarding flow table. This solution can realize a traffic balance on whole network links, improve the network throughput and improve performance of AI cluster service.