I have a simple network with two switches redundantly connected to each other and several computers on each switch talking back and forth across the two switches. The switches support RSTP/MSTP.
I want to be able to simulate a link loss between the two switches by disconnecting the active link and benchmark what the actual delay is until RSTP establishes a new topology and frame forwarding resumes but I'm not sure how to go about doing this.
I've tried measuring the Tx and Rx differential voltage to see if I can see a gap when no frames are being forwarded but it's exceptionally hard to trigger on such an event. Anyone have any ideas?
As a side note, the complexity of my network will grow quickly and I'm just trying to establish a good benchmark method from the start.