at a customer location we have a switching backbone running MSTP. The root switch (say SW A) has priority set to 4096. All vlans are in MST0. Two catalysts 6513 running rapid-pvst+ are connected to two switches of the Backbone (SW A and SW B). One of the 6513 has the default priority and the other has 40960. Now when i issue the command show spann vlan <x> on either of the 6513s i see that the root bridge is the catalyst with the default priority... as if the backbone wasn't there. More, the root catalyst has two ports to the backbone, and one of them is backup (blocked) as if the BB was a hub. I tested interworking MSTP-RSTP on my testbed using only catalysts and no problem. On the customer BB the switches are some other vendor. Is anyone aware of some interworking issues between Cisco (catalyst) and other vendors?
Thanks in advance
Here in fact, this seems to be an interoperability between PVST and a third party box. PVST BPDUs are sent to a Cisco specific multicast address that is not recognized by third party bridges. They are thus flooded through the core, which will indeed look like a hub. The only exception is vlan 1, that has its BPDUs sent to the standard IEEE mac address and that will interact with the third party bridges.
If the core was made of Cisco bridges, the behavior would be different. In order to avoid the flooding of the PVST+ BPDUs over the MST core, the Cisco MST switch would intercept them and try to interact with each individual PVST instance.
I'm guessing your question is whether MST works with 3rd party switch vendors?
I found a Cisco document while back that states "Most non-Cisco 802.1q switches maintain only a single instance of spanning tree, known as Mono Spanning Tree (MST)." You may want to verify which IEEE standards the 3rd party switch supports.