Application-Aware Elephant Flow Identification in High-Speed Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed data networks face challenges in efficiently identifying and managing large application-specific network traffic flows, particularly 'elephant flows' that consume significant bandwidth, leading to network congestion and performance issues.
Innovation Solution
The implementation of elephant flow identification and management logic within network devices, which filters and disaggregates network traffic based on application-specific criteria, uses multi-stage processes to identify and prioritize elephant flows, and manages their flow characteristics to mitigate impact on other traffic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional flow identification methods are used, then all network traffic is monitored, but resource requirements increase and false positives increase
Solution Approach 1:
The patent segments flow identification into two distinct stages: application-layer flow identification (ALFI) that identifies flows based on application-specific criteria, and traditional network-layer flow identification. This segmentation allows the system to focus computational resources on application-specific elephant flows rather than monitoring all network traffic, thereby reducing resource requirements while maintaining identification accuracy.
Solution Approach 2:
The patent introduces an application-aware flow identification module that acts as an intermediary between the application layer and network layer. This module uses application-specific information (such as application identifiers, port ranges, and protocol characteristics) to pre-filter and prioritize flows before they undergo traditional flow identification, reducing the burden on network devices and minimizing false positives.
2Reliability
If all network traffic is monitored for elephant flows, then comprehensive coverage is achieved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary flow identification at the application layer by matching packets against application-specific criteria (application identifiers, port ranges, protocol types) before full flow analysis. This preliminary action filters out non-candidate flows early in the processing pipeline, ensuring comprehensive coverage of potential elephant flows while significantly reducing the time and computational resources required for complete flow analysis.
3Measurement precision
If application-specific criteria are used for flow identification, then false positives are minimized, but implementation complexity increases
Solution Approach 1:
The patent implements a universal application-aware flow identification framework that can handle multiple application protocols and criteria through a single modular architecture. The system uses configurable application profiles that define identification rules for different applications, allowing the same infrastructure to adapt to various application-specific requirements without requiring separate implementation for each protocol, thereby managing complexity while maintaining high identification accuracy.
Data Source
AI summary
A network device identifies elephant flows. The network device filters received network data according to an application-specific criteria and identifies the elephant flow from the filtered network data. To do so, the network device can employ a multi-stage filtering process to identify an elephant flow in the received network data. The network device separates the filtered network data into multiple macroflows using a first hash function, and identifies the macroflow with the highest rate. Then, the network device disaggregates the high rate macroflow into multiple microflows using a second hash function and identifies the highest rate microflow as the elephant flow. The network device maintains an elephant flow cache with entries for currently identified elephant flows. The network device may also take management actions on the elephant flows, and the management actions may be application specific.