My CISCO Journal

BackboneFast:

In order to detect indirect link failure and to optimize network convergence time, Backbone Fast feature of STP is used. Backbone fast (BF, in short) is a CISCO proprietary feature. The term indirect link failure needs a little explanation. The link which is not directly connected to the core switch and which fails, such a link failure is called indirect failure. This indirect link failure is detected by a switch when it receives Inferior BPDUs! In order to understand Inferior and Superior BPDUs, we take following scenario:

Please note: f1/1 is in BLK and f1/2 is in FWD state

Suppose normal STP is running in our above topology. SW2 has been elected as our root bridge, BPDUs are continuously sent from SW2 to SW1 and SW3 every 2 seconds that SW2 has the lowest Bridge ID and it’s the root Bridge. SW1 has second lowest bridge ID. 

Now just imagine that the link between SW1 and SW2 goes down. As SW1 has second lowest bridge ID, and is now disconnected from SW2, it will proclaim itself as the root Bridge and will begin to advertise the same in its BPDUs, sending BPDUs to SW3, telling SW3 that it has the lowest bridge ID and it’s the root! At the same time SW3 is also receiving BPDUs from SW2, SW2 claims in its BPDUs to be the lowest in priority and the ultimate root bridge J Now to clear this confusion, SW3 compares both (SW1 and SW2) BPDUs, and it quickly realizes that BPDUs from SW1 are Inferior BPDUs and simply discards it. It only seriously considers the Superior BPDUs from SW2 only! Once Maxage Timer Expires on f1/1 port on SW3, it transitions into listening and after a certain time it begins to relay Superior BPDU data to SW1.

Now what role will backbonefast play, if it enables on all these switches? Backbonefast will minimize this Maxage timer interval. By enabling Backbone fast this Maxage stage is skipped, the delay is minimized from 50 seconds to 30 seconds! It sounds not a big deal but in a live network, such delay minimization at core switches greatly optimizes network performance.  All this magic is done by using Root Link Query protocol by switch once Backbonefast is enabled. Please remember one important thing, Backbonefast is always enabled on core switches, and to make all switches in a topology understand RLQ protocol, Backbonefast must be enabled on all switches in that topology!

The configuration of Backbonefast is quite simple. Its enabled globally by going into global configuration mode. The command to verify and configure Backbonefast is as follow:

Root Guard:

As the name suggests, in order to prevent entry of any new root switch into the network, Root Guard feature of STP is enabled on the interface to which new switch is going to be connected. Once Root Guard is enabled on an interface, it will discard all the superior BPDUs coming into that port and will change the port into Root-Inconsistent state; it will also discard Superior BPDUs until it stops receiving it.

Suppose in our above network topology we are going to connect a new switch to SW3 fa0/24. The Root Guard will be enabled as following on SW3 fa0/24:

If our new switch will send any Superior BPDU towards fast 0/24 of SW3, it will be discarded and port changed into Root-Inconsistent state until it stops such packets!

BPDU Guard:

In order to protect our network from loops, BPDU guard is configured on all ports on which Portfast is enabled. Because it’s expected that we can accidently plug any switch into our portfast enabled interface and can totally ruin our network by creating loops. Once BPDU Guard is enabled on an interface, it will discard any BPDU received and will instantly shutdown and will put the interface into err-disabled state.

To configure BPDU Guard on a specific interface, say SW1 fast 0/5 we use following commands:

To configure it on all ports, which by default must be running on portfast: