Stateful Network Traffic Simulation via Segmented Packet Processing

Solution Overview

Problem

Current high-speed network performance testing equipment is inadequate for simulating real-world scenarios in stateful network devices, as traditional stateless traffic generators fail to accurately measure the performance of sophisticated server load balancers and other stateful communications devices, which require full TCP/IP stack functionality.

Innovation Solution

A system utilizing programmable stateless packet processors and full TCP/IP stacks to simulate stateful TCP connections, with programmable stateless packet processors that react only to inbound packets, reducing hardware requirements and processing needs, while TCP amplification controllers modify test behavior based on performance metrics from stateful connections.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If traditional stateless packet blasters are used for network testing, then hardware requirements and costs are reduced, but the ability to accurately simulate real-world TCP session behavior and measure stateful device performance deteriorates

Engineering Contradiction:

Improvehardware requirementsVSAvoidsimulation accuracy

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The system divides the testing architecture into two distinct components: stateless packet processors that generate high-volume traffic at wire speed, and a separate full TCP/IP stack that simulates realistic TCP session behavior. This segmentation allows each component to specialize in its strength while working together to achieve both hardware efficiency and simulation accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The full TCP/IP stack acts as an intermediary between the stateless packet processors and the device under test. It receives packets from the packet processors, processes them through complete TCP session logic including state maintenance and feedback mechanisms, then forwards appropriately modified packets to the device under test, thereby enabling realistic simulation without requiring packet processors to maintain complex state.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If full TCP/IP stacks are implemented in all testing devices, then realistic TCP session simulation is achieved, but hardware requirements and processing needs increase

Engineering Contradiction:

Improvesimulation accuracyVSAvoidhardware requirements

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system divides the testing architecture into two distinct components: stateless packet processors that generate high-volume traffic at wire speed, and a separate full TCP/IP stack that simulates realistic TCP session behavior. This segmentation allows each component to specialize in its strength while working together to achieve both hardware efficiency and simulation accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of requiring every packet processor to implement complete TCP stacks, the system uses partial action by having dedicated full TCP/IP stack instances handle only the necessary session management for a limited number of controlled connections, while the majority of packet processing remains stateless and high-performance.

Inventive Principle:

Principle #16Partial or excessive action

3Speed

If stateless traffic generation is used, then processing speed is maintained, but the ability to model TCP feedback mechanisms and window size advertising deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidTCP session behavior modeling

Core Design Contradiction:

SpeedVSAdaptability or versatility

Solution Approach 1:

The system divides the testing architecture into two distinct components: stateless packet processors that generate high-volume traffic at wire speed, and a separate full TCP/IP stack that simulates realistic TCP session behavior. This segmentation allows each component to specialize in its strength while working together to achieve both hardware efficiency and simulation accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The full TCP/IP stack implements complete TCP feedback mechanisms including window size advertising, acknowledgment generation, and flow control. These feedback mechanisms dynamically adjust packet transmission based on simulated receiver performance, enabling realistic modeling of TCP session behavior while the stateless packet processors maintain high processing speed for the bulk traffic generation.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS8914432B2Real world traffic

Publication Date: 2014.12.16 KEYSIGHT TECH SINGAPORE (SALES) PTE LTD

AI summary

Methods and apparatus for testing real-world performance of an system under test are disclosed. The system under test is loaded with interactive transactions. A realistic mix of network traffic is simulated on the communications network. The performance of the system under test under load is tested. Accordingly, a real-world test of the system under test is provided.