Network Path Awareness Using Node-Specific Probe Packets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data center networks face complex path awareness solutions that require modifying packet structures and have high chip requirements, leading to increased packet length and complexity.
Innovation Solution
A path awareness method that sends request packets with node identifiers and reserved information, allowing for path determination without modifying packet structures, using simple solutions that reduce network overhead and fault troubleshooting time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If each intermediate node encapsulates its identifier into the probe packet, then the path can be tracked through multiple intermediate nodes, but the packet length becomes very long and the solution becomes complex
Solution Approach 1:
The path tracking function is segmented between different components: the source node segments the path into multiple probe packets, each targeting a specific intermediate node, while intermediate nodes segment the tracking task by responding only when they match the queried identifier. This avoids the need for a single complex packet to carry all intermediate node identifiers.
Solution Approach 2:
Instead of having intermediate nodes add their identifiers to the packet (forward direction), the invention inverts the approach by having the source node query each intermediate node's identifier separately and reconstruct the path from individual responses. This inversion simplifies the packet structure at each hop.
2Measurement precision
If probe packets are modified to include intermediate node identifiers, then path awareness can be achieved, but the requirement on chip becomes very high
Solution Approach 1:
Each intermediate node serves itself by responding to queries about its own identifier. The node simply checks if the queried identifier matches its own and responds accordingly, without needing complex packet modification capabilities. This self-service approach reduces chip requirements.
Solution Approach 2:
The source node creates multiple copies of probe packets, each with a different intermediate node identifier to query. Instead of modifying a single packet to include all identifiers, the system uses multiple simple packet copies, each targeting one intermediate node, thereby reducing individual packet complexity and chip requirements.
3Productivity
If a large quantity of components are used to implement distributed service, then the service scale can be large, but the coupling relationship between components becomes very complex
Solution Approach 1:
The path awareness function is segmented into independent probe packet transmissions to different intermediate nodes, rather than requiring a complex coordinated interaction among all components. This segmentation allows the system to scale to large numbers of components without increasing coupling complexity.
Solution Approach 2:
The same simple probe packet mechanism can be universally applied to query any intermediate node in the network, regardless of the total number of nodes. This universal approach allows the system to scale to large service sizes without requiring different complex coupling relationships for different scales.
Data Source
AI summary
This application provides a path awareness method and apparatus. The method includes: A source node sends, to a destination node, a first request packet that includes an identifier of a first node, an identifier of a first data flow, and first reserved information; receives a first response packet that includes the identifier of the first data flow, the identifier of the first node, and first indication information; and obtains, based on the first response packet, information indicating that a first path passes through the first node or not. The first request packet is used to request to determine whether the first path passes through the first node, the first reserved information is used to record a result about whether the first path passes through the first node.