What could possibly go wrong? I plug the cable in, get a link light, ping something and ensure there are no issues; I’m done, right? We’ve all done it a million times right?
I replaced two older switches with ones that support gigabit speeds. Followed the same routine that i mentioned earlier, the lights are green and tested fine with a couple of pings.
The other day I decided to do a Speedtest between two computers that traversed both switches. To my surprise, the speeds were less than 100 Mbps. So, I thought, why not document the troubleshooting steps so I could help others who might encounter the same issue.
I’ve said many times that you should baseline your network, AND your computer so you have some sort of reference for future testing. This was one example of that exact point. I knew that the computers could generate well over 100 Mbps of traffic so when I saw sub 100 Mbps speeds, I knew something was wrong.
The first step was to check the laptops’ network speed status and confirmed that they were both physically connected at 1 Gbps. I did this by verifying the adapter status under Network and Sharing Center->Adapter Settings->Adapter. Then I checked my switch to switch port status using the Cisco command “show interface status”, and noticed the uplink between them was 100 Mbps.
This is where the fun starts… I went over to physically inspect the cabling and realized the cables were quite old. The jacket was deteriorating and was probably there for over 20 years. I simply pulled a new run between the switches, tested with my NetAlly LinkRunner G2 and quickly confirmed that the new cable did in fact support one gigabit speeds. I then tested the existing cable with my LinkRunner G2 and confirmed that it was indeed only running at 100 Mbps, affirming one old adage of network troubleshooting, “Always suspect the physical layer first!”
Some people skip this step – validation – which drives me crazy. I’ve seen analysts replace a cable, check the speeds and say “Everything should be fine“ and then walk away. Please, take the extra few minutes to test and confirm that it “actually works as expected”. In some cases, something else may be ‘broke’ and cabling was only one part of the performance problem.
Lastly, ensure that you save some form of documentation of your test results. Fortunately, NetAlly testers can automatically upload the test data to their free cloud service, Link-Live. That way, I have a permanent record of every test – and can notate the results for future reference.
Iperf - https://iperf.fr/
NetAlly LinkRunner G2 - https://www.netally.com/products/linkrunnerg2/
The Technology Firm – https://www.thetechfirm.com