Some additional methods allow faster STP convergence if a link failure occurs.
Allows fast connectivity on ports by immediately switching to the Forwarding state without going through listening and learning. It is used with end devices and other devices that do not participate in Spanning-tree processes. Be careful with the command because it can cause loops and cause catastrophes with the network when a port configured with portfast is connected to another switch or bridge.
It is configured inside the ports:
Switch(config-if)# spanning-tree portfast
From global configuration, the following command assigns portfast to all commands that are in access mode
Switch(config)# spanning-tree portafast default
The following additional command is a macro that places the interface in access mode, enables portfast, and disables etherchannel
Switch(config-if)# switchport host
We can visualize the current status of PortFast with the following command:
Switch# show spanning-tree interface type member/module/number portfast
It allows a fast convergence of a blocked port to root if a physical failure is detected on the root port. That is, when the uplink of the primary root port fails, another blocked uplink can be immediately opened for use, avoiding the 50-second wait that STP uses to put a port in the forwarding state.
If more than one blocked port is a candidate for the root port, uplinkfast chooses the port with the best cost to the root bridge.
The uplinkfast configuration is global for the whole switch and all vlans:
Switch(config)# spanning-tree uplinkfast [max-update-rate pkts-per-second]
The uplinkfast command or function is not allowed on the root bridge. Recommended for use at the access layer as it raises the priority to 49,152, making it unlikely that the switch will be elected as a root bridge. The path cost of all ports on the local switch is increased by 3000, making the ports undesirable as routes to the root bridge of any downstream switch.
Switch# show spanning-tree uplinkfast
Allows fast convergence on a port that is blocking and must go to forwarding due to a failure on a neighbor’s root port. It saves the 20 seconds of hold time and goes straight to listening and then learning, shortening the convergence from 50 seconds to 30 seconds.
Switch(config)# spanning-tree backbonefast
By default, BackboneFast is disabled. When BackboneFast is used, it must be enabled on all switches in the network because BackboneFast requires the RLQ (Root Link Query) request and response mechanism to inform the switches of the stability of the root path. The RLQ protocol is active only when BackboneFast is enabled on a switch.
Check Backbone Fast:
Switch# show spanning-tree backbonefast BackboneFast is enabled Switch#