Cisco Systems 3560 Manual De Usuario

The IST is the only spanning-tree instance that sends and receives BPDUs. All of the other 
spanning-tree instance information is contained in M-records, which are encapsulated within MSTP 
BPDUs. Because the MSTP BPDU carries information for all instances, the number of BPDUs that 
need to be processed to support multiple spanning-tree instances is significantly reduced.

All MST instances within the same region share the same protocol timers, but each MST instance 
has its own topology parameters, such as root switch ID, root path cost, and so forth. By default, all 
VLANs are assigned to the IST. 

An MST instance is local to the region; for example, MST instance 1 in region A is independent of 
MST instance 1 in region B, even if regions A and B are interconnected. 

A common and internal spanning tree (CIST), which is a collection of the ISTs in each MST region, 
and the common spanning tree (CST) that interconnects the MST regions and single spanning trees. 

The spanning tree computed in a region appears as a subtree in the CST that encompasses the entire 
switched domain. The CIST is formed by the spanning-tree algorithm running among switches that 
support the IEEE 802.1w, IEEE 802.1s, and IEEE 802.1D standards. The CIST inside an MST 
region is the same as the CST outside a region.

For more information, see the 

 and the 

.

The implementation of the IEEE 802.1s standard, changes some of the terminology associated with MST 
implementations. For a summary of these changes, see 

.

The IST connects all the MSTP switches in a region. When the IST converges, the root of the IST 
becomes the CIST regional root (called the IST master before the implementation of the IEEE 802.1s 
standard) as shown in 

. It is the switch within the region with the lowest switch 

ID and path cost to the CIST root. The CIST regional root is also the CIST root if there is only one region 
in the network. If the CIST root is outside the region, one of the MSTP switches at the boundary of the 
region is selected as the CIST regional root.

When an MSTP switch initializes, it sends BPDUs claiming itself as the root of the CIST and the CIST 
regional root, with both of the path costs to the CIST root and to the CIST regional root set to zero. The 
switch also initializes all of its MST instances and claims to be the root for all of them. If the switch 
receives superior MST root information (lower switch ID, lower path cost, and so forth) than currently 
stored for the port, it relinquishes its claim as the CIST regional root. 

During initialization, a region might have many subregions, each with its own CIST regional root. As 
switches receive superior IST information, they leave their old subregions and join the new subregion 
that contains the true CIST regional root. Thus all subregions shrink, except for the one that contains the 
true CIST regional root. 

For correct operation, all switches in the MST region must agree on the same CIST regional root. 
Therefore, any two switches in the region only synchronize their port roles for an MST instance if they 
converge to a common CIST regional root. 

If there are multiple regions or legacy IEEE 802.1D switches within the network, MSTP establishes and 
maintains the CST, which includes all MST regions and all legacy STP switches in the network. The 
MST instances combine with the IST at the boundary of the region to become the CST.