SDN Data Path Delay Measurement via Controller-Injected Test Packets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In software-defined networks (SDNs), measuring data path performance between data plane elements is challenging due to increased latency and latency variation, which complicates network performance measurement and administration.
Innovation Solution
A method is implemented in the network controller to send control messages with classification rules to ingress and egress network elements, modifying data packets with packet identifiers and timestamps, allowing for the calculation of delay between these elements based on received timestamps, thereby enabling accurate performance measurement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data path performance measurement is implemented in SDN, then network performance monitoring capability is improved, but device complexity and measurement difficulty increase
Solution Approach 1:
The patent introduces an intermediary measurement mechanism where the network controller acts as a mediator between data plane elements. The controller injects test packets with unique identifiers into the data path and collects response packets from data plane elements, enabling performance measurement without directly modifying the data plane devices themselves. This intermediary approach simplifies the complexity at data plane elements while maintaining measurement capability at the control plane.
Solution Approach 2:
The patent uses packet copying where test packets are created as copies with modified headers (including unique identifiers and timestamps) and injected into the data path. These copied packets traverse the data path alongside normal traffic but are distinguished by their measurement-specific headers. The response packets are also copies that return to the controller with measurement data, enabling non-intrusive performance measurement.
2Measurement precision
If packet modification with identifiers is performed, then measurement accuracy is improved, but processing overhead increases
Solution Approach 1:
The patent applies partial action by only modifying specific packets that are designated for measurement purposes, rather than all packets in the network. The network controller selectively injects test packets with measurement headers into the data path, leaving normal data packets unmodified. This partial modification approach minimizes processing overhead while maintaining measurement accuracy for the selected test packets.
Solution Approach 2:
The patent implements preliminary action by pre-configuring measurement parameters, packet identifiers, and classification rules in the network controller before actual measurement begins. The controller prepares measurement templates and criteria in advance, so that when test packets are injected, the modification and measurement processes can proceed efficiently without real-time configuration overhead.
3Measurement precision
If classification rules are implemented for packet selection, then measurement specificity is improved, but control message complexity increases
Solution Approach 1:
The patent segments the measurement process into distinct phases: packet selection based on classification rules, packet injection with identifiers, traversal through data path, and response collection. Each phase has its own simplified control messages and processing logic. The classification rules themselves are segmented into specific criteria (packet type, source/destination, protocol) that can be independently configured and evaluated, reducing the complexity of any single control message while maintaining overall measurement specificity.
Data Source
AI summary
A method in a network controller of a control plane in a software defined network (SDN) coupled to a plurality of network elements (NEs) of a data plane in the SDN is described. The method includes sending a first control message having content including a first set of one or more classification rules to an ingress NE of the plurality of NEs; sending a second control message having content including a second set of one or more classification rules to an egress NE of the plurality of NEs; receiving the second notification message from the egress NE responsive to the selected data packet having been received by the ingress NE from an upstream NE and forwarded through the SDN from the ingress NE toward the egress NE; and calculating an indication of a delay of the selected data packet between the ingress NE and the egress NE.