Unusual spanning tree configuration

[pcobley]

Hi,

We are migrating 2 old sites to one new larger site. During the migration stage the 'old' core switches will be connected to the 'new' core with redundant links.

The old network is flat with all ports in vlan 1.

The two sites are in separate VTP domains and we are creating new vlan's in the new building.

Devices in the old network's vlan1 are in the same subnet range as the new network's vlan20. To remove the vlan information from the header and allow the network to function correctly, i am connecting vlan 1 access ports in the old building to vlan 20 access ports in the new building.

Can someone please confirm that spanning tree will succesfully block one of these ports to break the physical loop, even though the ports are in different vlan's.

Thanks in advance,

Paul.

Summary of connections :-

new switch A port 1/1 trunk link to new switch B port 1/1

new switch A port 2/1 vlan20 access port link to old switch A port 2/1 vlan1 access port

new switch B port 2/1 vlan20 access port link to old switch B port 2/1 vlan1 access port

old switch A port 1/1 trunk link to old switch B port 1/1

[milan.kulik]

Yes, this should work.

Just be carfull regarding the STP root: there will be one common STP tree including "old VLAN1" and "new VLAN20" so there will be only one root bridge in this tree. So in one of these parts the root brige will be changed. And depending on the new root and line speeds there will one line be blocked (it might even be the old trunk or VLAN20 on the new trunk, not the interconnecting access line).

Regards,

Milan