No. Unfortunately there is no such mechanism defined by the MST spec. The manual configuration of the mst regions is actually the trick that makes the protocol relatively simple... but not very flexible I admit.
If you are running CatOS however there is a way to achieve this: VTP3 will take care of propagating the MST configuration along the VTP domain.
Else, you'll have to configure the switches one by one, which is going to be painful and will disturb your traffic.
You can take this opportunity to plan a little bit for the future. What you can do is create in the MST configuration more instances than you need and map evenly all the vlans to them.
Suppose you currently need 2 instances and 10 vlans and that you may require up to 10 instances in the future:
- create 10 instances by mapping 400 vlans to each (well, of course map the 10 vlans you currently use to the two instances you currently need).
- make sure that the vlans for the 8 instances you don't currently need are not created (here I mean not created in the VTP database).
Our MST code is optimized so that an instance is not run unless there is at least one vlan that is mapped to it and that is active on one port in the switch (the only exception is instance 0 that always exists.)
Thus, the current setup does not require more resources that if you only had configured two instances with 10 vlans. Now, if you need a new vlan in one of your two current instances, you have plenty to choose from. If you need some vlans with a different topology: just use one that is mapped to an usued instance. This way, you can at least quietly add/remove vlan without disrupting the traffic because you are not modifying the VTP configuration.
Hope this helps,
Francois
