Home » Layer-2 Switching » MSTP (802.1S)

MSTP (802.1S)

Multiple Spanning Tree Protocol

With Per Vlan Spanning Tree protocol, you can have a different spanning tree topology per Vlan, with Root Switch for one vlan on a switch and Root Switch for another vlan on another switch. If you have a large number of vlans then each of the vlan can have its own topology. This can also become too complicated with large number of vlans and many switches, so to simplify the topology and make them more administratively manageable MSTP was introduced where you can now define the logical topologies you want in the network and let the Vlans  you choose traverse over the defined topologies. In MSTP the logical topology is called an instance.

Example:

In the above diagram suppose you have vlans 2-10 then with MSTP, you can take Vlans 2,3,4 and map them to MSTP Instance 1 and map Vlans 5,6,7 to Instance 2 and the remaining Vlans 8,9,10 to MSTP instance-3. Overall result is that you are engineering the traffic for multiple vlans on different logical topologies using MSTP.

In instance 1 the links between SW2 and SW4 are blocked
In Instance 2 the links between SW1 and SW3 are blocked
In instance 3 the links between SW3 and SW4 are blocked

An MSTP Region has to be created in order to run the MSTP.
MSTP Region is a collection of switches which have the same view of the underlying Physical topology. When MSTP is a configured a special internal  instance is automatically created and is called as Instance 0.
Instance 0 is used to carry all Spanning Tree information for the entire physical topology and all vlans are mapped to this instance 0 automatically.

Note: MSTP does not send out separate BPDUs per logical topology instances.
The Spanning Tree required information is carried by the BPDUs in this Special internal instance using M Record fields – which is a Type-Length-Value (TLV) field.
Note: MST Instance 0 is also called as Internal Spanning Tree (IST) and  it will carry all the STP needed information for all vlans, and since its there by default on all the switches and carrying the BPDUs for all vlans, any misconfiguration of van mapping can be detected.

Note: Hello, Fwd-Delay and Max-age Timers cannot be configured differently for each MST instance separately. Each MST instance inherits the timers from the IST instance. IF you need to change the MST timers, then it has to be done on the IST instance.

Remaining Hops BPDU Field
In MSTP, the BPDU has a field called Remaining Hops, this special BPDU field is used to age out old information by decrementing the hop count of the BPDU at every downstream switch, so the BPDU does not flow endlessly in the network. The Max Hops value for the BPDU can be configured on the switches.

STP Dispute Feature of MSTP
This is similar to the STP Loop Guard in one particular situation explained below
In a Situation where an uplink Switch sends superior root switch information to the downstream switch, and in return the downstream switch sends an inferior root switch BPDU with the Designated bit set, it means that the downstream switch is not able to listen to the BPDUs from upstream switch. In this case the Upstream switch will block the port connected to the downstream switch effectively breaking the loop. Still Loop Guard must be configured in the MSTP switches to save the topology from loops created due to all other reasons.

Configuring MSTP
Steps:
1. Define the Spanning Tree mode as MST
2. Enter into MST configuration mode
3. Give a name to MST region
4. define a revision number for MST (this is not the VTP revision number)
5. Create MST instances and map the vlans for each instance ( optional: If no instances are defined then all vlans will be mapped to instance 0 by default)
6. Define Root switches for each of the instances (optional)

An MSTP region is a collection of switches which share the common view of the physical network topology. For switches to become the member of that region they should have a match on the following
1. MSTP region name
2. Config Rev Number
3. VLANs to Instance mapping

In the below example we will configure
vlans 2,3,4 in instance 1
vlans 5,6,7 in instance 2
vlans 8,9,10  in instance 3

Note: All other remaining vlans are automatically mapped to the instance 0 which is created by default in MSTP.

Commands on all 4 switches to configure MSTP:

spanning-tree mode mst
!
spanning-tree mst configuration
name region3
revision 1
instance 1 vlan 2,3,4
instance 2 vlan 5,6,7
instance 3 vlan 8,9,10
!
end

Setting Root switches in MST.
You can define the a root switch for each instance in MSTP and the switch will become root for all the vlans in that instance.

SW1# spanning-tree mst 1 root primary  ( SW1 will be root for all vlans in instance 1)
SW2# spanning-tree mst 2 root primary  ( SW2 will be root for all vlans in instance 2)
SW3# spanning-tree mst 3 root primary  ( SW3 will be root for all vlans in instance 3)

OR you can use the Priority command which is

SW1# spanning-tree mst 1 priority 4096

SW2# spanning-tree mst 2 priority 4096
SW3# spanning-tree mst 3 priority 4096

Verify MSTP
# show spanning-tree mst configuration
# show spanning-tree mst
# show span mst detail
# show spanning-tree mst <instance number>
# show spanning-tree mst interface <type/num>


Incoming search terms for the article:

Leave a Reply