Autonomous Flow Statistics Relay in SDN Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In software-defined networking (SDN) environments, the current approach of periodic statistics requests from the network controller to data plane devices leads to high bandwidth requirements and increased processing loads, especially in large-scale deployments, and lacks an autonomous alerting mechanism for detecting bursty or elephant flows without explicit requests.
Innovation Solution
Implementing statistics profiles on network devices that allow them to autonomously relay flow statistics when specific criteria are met, reducing the need for continuous requests and enabling batched alerts, thereby reducing bandwidth usage and processing loads in the network controller.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If periodic statistics requests are sent from the network controller to data plane devices, then real-time flow statistics can be collected, but bandwidth usage on the control network increases enormously
Solution Approach 1:
Instead of the network controller periodically requesting statistics from data plane devices, the patent inverts the approach by having data plane devices autonomously generate and send statistics messages when flow criteria are met. This reversal eliminates the need for continuous control plane requests while maintaining effective statistics collection.
Solution Approach 2:
Data plane devices are empowered to autonomously monitor their own flow statistics and self-determine when to report to the network controller based on predefined criteria. This self-service mechanism eliminates dependency on periodic controller requests, reducing control network bandwidth consumption while ensuring timely statistics reporting.
2Measurement precision
If periodic statistics requests are sent from the network controller, then flow statistics can be gathered, but processing load on the network controller increases
Solution Approach 1:
Data plane devices autonomously evaluate flow criteria and generate statistics messages without requiring controller intervention. This shifts the processing burden from the network controller to the data plane devices, reducing controller complexity while maintaining effective statistics collection.
Solution Approach 2:
Flow criteria are pre-configured on data plane devices, enabling them to autonomously evaluate incoming flows against these criteria and generate statistics messages when conditions are met. This preliminary configuration eliminates the need for continuous controller processing while ensuring accurate statistics collection.
3Speed
If continuous requests for statistics are transmitted, then real-time network state information is obtained, but control network bandwidth is consumed excessively
Solution Approach 1:
Instead of continuous periodic requests from the controller, statistics messages are transmitted event-driven when flow criteria are satisfied. This transforms continuous bandwidth consumption into targeted, criterion-based transmissions, maintaining real-time monitoring capability while dramatically reducing control network bandwidth usage.
Solution Approach 2:
Data plane devices autonomously determine when to transmit statistics based on flow criteria evaluation, eliminating the need for continuous controller-initiated requests. This self-service approach maintains timely statistics retrieval while optimizing bandwidth consumption by transmitting only when necessary.
4Measurement precision
If the control plane continuously monitors flow statistics, then network anomalies can be detected, but the system lacks autonomous alerting capability
Solution Approach 1:
Data plane devices autonomously evaluate flow criteria and self-determine when to generate statistics messages, providing built-in autonomous alerting capability. This eliminates the need for continuous control plane monitoring while maintaining effective anomaly detection through criterion-based autonomous reporting.
Solution Approach 2:
The system implements criterion-based feedback where data plane devices automatically generate statistics messages when predefined flow conditions are met. This feedback mechanism provides autonomous alerting for network anomalies without requiring continuous controller intervention, enhancing automation while maintaining detection precision.
Data Source
AI summary
Methods and apparatuses for autonomous relay of statistics from a network device to a network controller of a software-defined networking (SDN) network are described. The network controller is operative to transmit a message including a statistics profile, where the statistics profile includes a type of data to be collected associated with a behavior of a flow at the network device and a set of one or more criteria; where the message causes the network device to transmit a statistics message associated with one or more flows when the one or more flows satisfy at least one criterion from the set of one or more criteria, where the statistics message includes data collected at the network device associated with a behavior of the one or more flows as identified by the type of data to be collected included in the statistics profile.