Hello khan,

I try to give you some info.

IEEE STP 802.1D is mono-instance : only one STP manages the topology redundancy for all Vlans defined in the switched networks this lead to the use of a single topology for all Vlans: links placed in blocking state are empty while links in forwarding state can experience high link usage.

PVST+ is supported since CatOS 4.1 so it is unlike you will deal with its parent PVST.

PVST+ uses a different approach: for each Vlan that exists in the switched network a separate STP instance runs : it sends its own BPDUs and receive them.

So with PVST+ more the one topology is possible: STP cost can be changed on links on a per Vlan basis and a different switch can be elected as the route bridge this allows some form of load balancing over L2 trunk links: the same L2 trunk link can be in forwarding state for Vlan 10 and blocked for vlan 20, another link can be in forwardng for Vlan 20 and blocking for Vlan 10

In this way some load-balancing is possible.

The limit of PVST+ is scalability in a MAN scenario.

Besides this most enterprise design with an access layer and a distribution layer need only two topologies : one with route bridge distribution1 and one with root bridge distribution2.

MST tries to overcome this limitation : instead of a single instance or a 1 to 1 relation between Vlans and STP instances it provides the capability to create some instances (up to 16) and to map the Vlans to each instance.

In this way scalability is achieved leaving space for some form of load-balancing.

Each MST instance is a Rapid STP instance 802.1W.

MST require the configuration of a region and within region config the vlans are mapped to the MST instances.

see:

http://www.cisco.com/en/US/docs/switches/lan/catalyst6500/ios/12.2SX/configuration/guide/spantree.html

Hope to help

Giuseppe