Spanning Tree Protocol - STP

 

Switches and Bridges use STP to prevent bridging loops.  It is actually a fairly complex process, so the we will only discuss an overview here.

 

First, two timers are decided upon – a “Hello” timer, and a “MaxAge” timer.  Then, all devices are originally assigned as a root, until they collectively decide which one is the root.  Then the root begins sending out exploratory CBPDU’s (Configuration Bridge Protocol Data Units) to the devices – one every “hello time”.  The unit with the lowest cost path to the root unit becomes a “designated” or “root port” unit and is placed in the forwarding state. 

 

  1. the root sends a CBPDU, cost 0 out all it’s interfaces – this is repeated every Hello time.  The default Hello time is 2 seconds

  2. the bridges connected to the root add their cost and send CBPDU’s out their other ports (not the port connected to the root)

  3. repeat step 2 to each bridge in the network – they learn the lowest cost paths to the other bridges, so that each bridge will use just one path to each other bridge

 

The proccess continue forever – if a bridge does not receive a CBPDU within the MaxAge time, it restarts the spanning tree process with other bridges

 

There are 4 states (2 are Stable, and 2 are Transitionary) :

 

Stable States -  Blocking and Forwarding

Transitionary States – Listening and Learning