One of the biggest mistakes I see out there when I observe analysts conducting performance tests is that they get paralyzed by trying to include way too much detail upfront. During this exercise, they feel so overwhelmed that they don’t bother testing since they realize there are so many variables to document.
When I train or present, I say, 'Just start measuring.' If you see an anomaly or something out of place, then start digging in and documenting. For example, if my Wi-Fi connection was rated at 800+ Mbps and my testing reported 11 Mbps, I would look into my access point configuration, like channel selection, channel width, and more. If the equipment configuration was fine, I would then use a Wi-Fi or RF spectrum analyzer in an effort to determine the root cause, like RF interference.
On the other side of the coin, if your performance aligns with your expectations, you now have a baseline or snapshot of the current performance to refer to in the future. I would say that about 75% of the time, clients do not have a baseline, trace files, or current documentation.
In this example, I am comparing iPerf3 on various devices and network topologies to illustrate that the results will always be different.
See other troubleshooting and testing videos by Tony Fortunato here.
Related articles:
Discover the benefits of scalable backbone and network services for application and cloud providers.
Network automation is gaining momentum. Here's how you can drive this powerful technology to its full potential.
Subsea cable alternatives can provide a level of comfort for those concerned about disruptions. Realistically, they will continue to play a small, but critical role in the short term.
