I have a really strange loopback problem..
Sometimes, when my network topology change (like a link goes down), about half of my 802.1q trunks over the network goes err-disabled because of a loopback detection.
I've triple-checked... There's absolutly no loopback not managed by STP on my network.
I really don't understand why only half of my switches are detecting the problems and others don't...
I have about 120 switches on the network (80X2950 (12.1(20)), 30X2960 (12.2(25)), 10X3750 (12.2(20)))
To patch the problem, I was considering stopping the errdisable for the loopback cause or at least, making a low recovery interval for this cause...
Someone have any advices to give me??