[c-nsp] Maximum spannig tree instances

[Saku Ytti]

On (2009-07-17 15:55 +0100), Phil Mayers wrote:

> In addition, I have serious concerns about the scope of instance 0,
> particularly in the topology we run (a collapsed core/distribution
> triangle).
> 
> When I tried it on the bench, I could not come up with an MST setup
> that worked.

If working is setup where each host in L2 has connection to other hosts
in the L2, then MST really should work in any topology where classical
STP would work. So assuming you'd dump all VLANs to single user
MST instance, it would be like classical STP, but with faster
convergency.

MST/PVST boundary operation can cause lot of damage, unless carefully
planned. But as always, legacy networks and migrations are the hardest.
I'd have juicy example what can happen when you have PVST and MST access
networks and you start selling EoMPLS.
L2 ethernet is so wonderful way to break lot fast.


-- 
  ++ytti