Sorry about the hand written drawing. I am away for a funeral and do not have Visio on my laptop. We have 2 6509's connected via VSS. We have 3 different campus buildings with 2 switch stacks is each floor closet. The date center has around 45 top of rack switch's uplinked the same as closets with a dual 1gb port channel links. The STP issue happens on any port channel pair when both links have gone down. An example of this is when we have storms a closet stack has a UPS but could drain and drop power to the switch stacks for that floor. When they come up the firt 1gb port to come up goes through spanning tree blocking, then forwarding. When it switches to forwarding something starts that spikes the core cpu, core CLI including serial console is stuck. It relives in 6-10 seconds. It does affect packet forwarding network wide as it is the core. I can replicate it at will by downing a pair of uplinks from the core side and when I bring the first port up for the port channel it does it every time when STP moves to forwarding. I have not tried any debug yet as every time I do this it causes impact, alerts, etc. Again, if I add the config line Spanning-tree bpdufilter enable to the port on the 6509 I bring up there is no impact.
