Spanning Tree Protocol STP Tutorial

To build fault tolerance into a network, it’s common to connect switches with redundant paths. But redundant paths between the same two segments create a side effect: packets can end up looping endlessly back and forth across those paths. That condition is known as a bridging loop.

Note: throughout this tutorial, “bridge” and “switch” are used interchangeably, as is typical when discussing STP.

To stop bridging loops from happening, the IEEE 802.1D committee defined the Spanning Tree Algorithm (STA), commonly known as the Spanning Tree Protocol (STP). STP is a link-management protocol: it keeps the redundancy those extra paths provide while actively preventing the loops that would otherwise result. For an Ethernet network to work correctly, only one active path can exist between any two stations at a time.

What happens without loop prevention

Picture two switches connected by two redundant links, with PC A attached to one switch and PC B attached to the other. PC A wants to talk to PC B, so it sends out a broadcast — say, an ARP request — to find PC B’s location.