The 802.1w standard or Rapid Spanning Tree Protocol (RSTP) was defined to make some improvements over STP. The main problem introduced with STP is the amount of time it may take for a new configuration to be established when there are changes to the topology. During this transition period networks may be isolated and packets can be dropped. RSTP was designed to address this issue and reduce the time it takes for topology changes to occur.

STP defines five different port states as outlined earlier in the chapter. Since a port that is in a blocking, disabled, or listening state does not forward any frames, RSTP combines these into a single port state known as a Discarding port state as listed below:

802.1d STP Port States 802.1w RSTP Port States
Disabled Discarding
Blocking Discarding
Listening Discarding
Learning Learning
Forwarding Forwarding

RSTP also introduces a new concept referred to as port roles. Along with the two port roles of root port and designated port, RSTP uses alternate ports and backup ports. These two new port roles correspond to the blocking state de-scribed in the 802.1d standard. An alternate port is one that is blocked from receiving more useful BDPUs from a different switch. A backup port is one that is blocked from receiving more useful BDPUs from the same switch it is on.

With 802.1d, a non-root switch will only generate a BDPU if it receives one on its root port. With 802.1w, switches now sent out their own BDPUs every hello time regardless of whether a BDPU has been received on the root port or not. The benefit of this is that the BDPUs can now be used as a keep alive mechanism between switches. So if a neighboring switch does not respond to three consecutive hello messages, it is assumed that connectivity has been lost.

The hello time is the length of time between BDPU transmissions. The default value is 2 seconds but can be set to a value between 1 and 10 seconds.

RSTP was designed to decrease the amount of time for a new topology to be agreed upon. Convergence speed in increased in the following ways:

  • Faster Aging – A connection is considered to be lost if hellos are not received for three consecutive times (by default this will be six seconds). This means that failures are detected much faster.
  • Rapid Transition – When a topology change is received from the root, a switch can immediately decide upon a new root port without having to wait for a predefined time interval to elapse.
  • Fast Backup – A backup port is immediately changed to a for-warding state when a switch loses its root port.

[awsbullet:spanning+tree+protocol]