I will start from the beginning. A client was complaining about performance issues. Specifically, the symptom was that a computer seemed to freeze or stutter at random times. During the initial meeting, I found out that this issue has been around since day one and that all the computers are based off a ‘standard image.’
This is when my standard workflow pays off. I treated this like any other troubleshooting engagement. The first step is to get a boot up and login baseline. The boot up and login will identify the protocols in use as well as the servers that the client communicates with. With this information, I may be able to identify some areas for configuration optimization. I used a ProfiShark 1G tap that was USB connected to my computer.
In this case, both login and boot up traces did not reveal any anomalies. Then I captured 5 minutes of packets while the client’s computer was idle. The key here is to close all obvious applications and leave the computer alone. This will identify the protocols and servers the client communicates with during regular operation.
I’m glad I took the 5-minute idle trace because I found a strange conversation between the client’s computer and a Sony Bravia smart TV. After further testing, we were able to correlate this conversation with the computer performance issues eight out of ten times. The other noteworthy point is that within the client packets we saw a reference to Spotify. After further investigation, the client acknowledged that he installed the Spotify client after getting admin access to his computer for a project.
We uninstalled the Spotify software and disconnected the smart TV from the network until that switch port VLAN configuration was modified. Watch the entire troubleshooting process here: