Microburst Testing via Segmented Packet Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network testing methods fail to effectively detect and analyze microbursts, which are short spikes of packets that can cause packet loss and delays, as they are often overlooked due to averaging over long durations, leading to incomplete performance metrics and potential issues with buffer allocation.
Innovation Solution
A method and system for microburst testing that divides tests into cycles with equal packet distribution across ports, utilizing replication counts to simulate both microburst and non-microburst phases, allowing for monitoring of buffering capacity and identification of peak loads without packet loss, using a test platform that sends packets at rates exceeding the network switch's processing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If average traffic statistics are monitored over long durations, then measurement stability is improved, but microburst detection capability deteriorates
Solution Approach 1:
The patent segments the monitoring period into multiple short time intervals (e.g., 1 second intervals) within a longer measurement window. Instead of computing a single average over the entire duration, the system calculates separate averages for each short interval, enabling detection of brief traffic spikes while maintaining overall measurement stability through the structured approach.
Solution Approach 2:
The system implements periodic monitoring by dividing the measurement period into regular, discrete time intervals and computing statistics for each interval separately. This periodic structure allows the system to capture transient microburst events that occur within specific intervals while maintaining stable long-term measurement through the regular sampling pattern.
2Measurement precision
If packet transmission rate exceeds processing speed to simulate microburst, then microburst condition accuracy is improved, but packet loss increases
Solution Approach 1:
The system performs preliminary actions by pre-configuring buffer allocation and test parameters before initiating the microburst simulation. The test platform is prepared with appropriate buffer sizes and packet generation rates configured in advance, allowing the system to handle the expected packet overflow conditions without actual data loss, as the buffers are pre-allocated to accommodate the simulated microburst.
Solution Approach 2:
The system uses virtual copies and simulation rather than actual packet transmission that would cause permanent loss. The microburst condition is replicated through virtual packet generation and buffer management, allowing the test to accurately simulate overflow conditions while the actual data can be regenerated or recovered through retransmission protocols.
3Productivity
If buffers are used to queue packets during microburst, then packet handling capability is improved, but buffer overflow risk increases
Solution Approach 1:
The system implements feedback mechanisms by continuously monitoring buffer utilization levels during packet queuing. When buffer usage approaches capacity thresholds, the system adjusts packet generation rates or triggers flow control measures, creating a feedback loop that prevents buffer overflow while maintaining optimal packet handling capability during microburst conditions.
Data Source
AI summary
Methods, systems, and computer readable media for microburst testing are disclosed. According to one method, the method includes defining a test that is divided into one or more test cycles, wherein each test cycle includes a plurality of time intervals for providing packets to destination ports of N ports of the network equipment test device via a plurality of ports of a network switch under test, wherein N is an integer. The method further includes initializing a replication count. The method also includes executing an instance of the test using the replication count. The method further includes monitoring packets switched by each port of the network switch under test to determine an indication of buffering capacity of each port of the network switch under test.