This post is a short one, but hopefully still valuable. I spent 30 minutes troubleshooting a flapping switch port in the lab this week. I was attaching a headless test device to a 3750G and I couldn’t see the status of the NIC from the tester’s perspective.
So one crash cart and a bit of port-swapping later, I found the problem. It was a dodgy switch port and swapping to another port (with identicial config) solved the issue. Then it hit me. I had hit this particular bad port before but didn’t follow up. This is network Karma, and I deserved that wasted 30 minutes.
So what next? The ‘right’ thing to do here is to RMA the switch, but I have plenty of ports in the switch stack and the RMA isn’t a priority. I’ve configured the port using ‘description ** Faulty Port **‘, but that’s not enough.
Then I remembered a bag of RJ-45 connectors I had in the cabinet, and inserted a single connector to occupy our dead port, port 12.
This is a quick fix but should make it harder for the next admin (or me) to re-use that bad port.