The Story of Spanning Tree Protocol (STP)

The Begining of the Story

#

Back in the early days of Ethernet, redundancy was necessary. If one cable or switch failed → another path needed to take over.

👉 To achieve this, engineers connected switches in loops. Great idea… until it wasn’t.

🚨 The Problem – Loops Gone Wild

#

When redundant links formed loops, chaos followed:

1. Broadcast Storms → Broadcast frames (like ARP requests) endlessly circled → consuming bandwidth & CPU.
2. MAC Table Instability → Switches kept updating MAC tables incorrectly → same frame arriving from different ports.
3. Duplicate Frames → Devices received multiple copies → inefficiency + confusion.

Networks basically choked themselves.

💡 The Solution – A Tree, Not a Mess

#

Radia Perlman (often called the “Mother of the Internet”) designed Spanning Tree Protocol (STP).

→ STP detects loops → blocks redundant paths → reactivates them only if needed. → You can design networks with redundancy but still stay loop-free.

How Does STP Work?

#

When switches boot up on a redundant network:

1. Each switch assumes it is the Root Bridge 👑
2. It starts sending BPDUs (Bridge Protocol Data Units) every 2 seconds → “Hey, I’m the root!”

🏆 Root Bridge Election

#

Switches compare Bridge IDs to decide who’s the real Root Bridge.

Bridge ID = Bridge Priority + MAC Address

• Default Bridge Priority = 32768 (range 0 – 65535)
• If priorities are the same → switch with the lowest MAC address wins 🥇

Once a switch hears a better BPDU → it gives up and forwards that BPDU instead.

🌉 Root Port Election

#

After the Root Bridge is chosen → all other switches pick their Root Port.

• The Root Port = port with the lowest cost path to reach the Root Bridge.
• STP path costs depend on link speed:

👉 Faster links = lower cost.

📍 Designated & Blocking Ports

#

To break loops:

• Each network segment (collision domain) must have one Designated Port → the port with lowest cost to the Root Bridge.
• All other redundant ports → go into Blocking State.
• Blocking Ports do not forward traffic → preventing loops.

🔄 STP Convergence

#

STP transitions ports through states before forwarding traffic:

👉 Total recovery time from a failure = 50 seconds. That’s a lifetime in networking.

⚡ Failures & Recovery

#

Example: If a Root Port fails →

1. Switch notices no more BPDUs.
2. Blocked port moves → Listening → Learning → Forwarding.
3. Connectivity restored… after 50 seconds ⏳

💡 Modern networks use RSTP (Rapid Spanning Tree Protocol) → recovery in 1–2 seconds.

🧩 BPDU Basics

#

Each BPDU carries:

1. Bridge ID (BID) = Priority + MAC Address
2. Root Bridge ID = Who the root is (initially, itself)
3. Path Cost = Total cost to reach the root

Cisco Configurations

#

STP is on by default in Cisco switches. But here are useful tweaks:

Enable STP (default PVST):

Switch(config)# spanning-tree mode pvst

Set a switch as Root Bridge (lower priority wins):

Switch(config)# spanning-tree vlan 10 priority 4096

Force port into specific STP role:

Switch(config-if)# spanning-tree port-priority 64

Enable RSTP for faster convergence:

Switch(config)# spanning-tree mode rapid-pvst

Limitations – No Load Balancing

#

STP builds a single active path → no real load balancing.

If multiple equal-cost paths exist → it simply picks the neighbor with the lowest Bridge ID. The rest sit idle in blocking state.

🌳 Wrapping Up

#

Without STP, networks with redundant links would collapse under their own loops.

• Redundancy is good → but needs control.
• STP keeps one tree-like active topology.
• RSTP makes it much faster in modern networks.

Every time you connect redundant switches, just remember → Radia Perlman saved the day with a tree. 🌲