CHAPTER 07
Beginner
Link State Routing Protocols
Updated: May 15, 2026
20 min read
# CHAPTER 7
Link State Routing Protocols
1. Introduction
Distance Vector protocols (like RIP) failed because they routed by rumor; routers never possessed a true map of the network. As enterprise networks grew to encompass thousands of routers, engineers needed a protocol that was blindingly fast, fiercely intelligent, and completely aware of the physical topology. The answer was the Link-State algorithm. The undisputed king of modern enterprise Link-State routing is OSPF (Open Shortest Path First). In this chapter, we will explore the immense mathematical power of OSPF, understand how it uses Link-State Advertisements (LSAs) to build a global map, and see how the famous SPF algorithm calculates the perfect path based on actual bandwidth speed.2. Learning Objectives
By the end of this chapter, you will be able to:- Contrast Link-State routing (Topological map) with Distance Vector routing (Rumor).
- Define Link-State Advertisements (LSAs) and how they populate the database.
- Explain the function of the Shortest Path First (SPF / Dijkstra) algorithm.
- Understand why OSPF uses Bandwidth (Cost) as its primary routing metric.
- Define the hierarchical concept of OSPF Areas (Area 0).
3. Beginner-friendly Explanations
The GPS Satellite Map (Link-State): Remember how RIP was just listening to rumors from the router next door? OSPF works like a drone scanning the entire city.- 1. Every router in the building acts like a reporter.
- 2. Router A broadcasts a message to the *entire building*: *"I am Router A. I have a 1 Gigabit Fiber cable connected to Router B."*
- 3. Router B broadcasts: *"I am Router B. I have a slow DSL cable connected to Router C."*
- 4. Every single router collects all these reports and builds a complete, identical, physical map of the entire building in its memory.
- 5. Because every router has the exact same map, they independently run a mathematical algorithm to find the absolute fastest path to the destination.
4. Link-State Advertisements (LSAs)
The "reports" that routers broadcast are called LSAs (Link-State Advertisements). Unlike RIP, which sends its entire routing table every 30 seconds (wasting massive amounts of bandwidth), OSPF only sends an LSA when something actually changes. If a cable breaks, the router attached to it instantly fires an LSA across the network. Because the updates are instantaneous and small, OSPF achieves lightning-fast Convergence.5. The SPF Algorithm (Dijkstra's Math)
Once a router has collected all the LSAs and built the map (the Link-State Database), it runs the Shortest Path First (SPF) Algorithm (invented by computer scientist Edsger Dijkstra).The Metric: Cost (Bandwidth) OSPF is not blind like RIP. OSPF looks at the physical speed of the cable.
- A fast 10-Gigabit cable is given a Cost of 1.
- A slow 10-Megabit cable is given a Cost of 100.
6. OSPF Hierarchy: Areas
Building a mathematical map of 5 routers is easy. What if a global bank has 5,000 routers? If one cable breaks in London, you don't want 5,000 routers worldwide to recalculate their SPF algorithm simultaneously; the CPUs would melt.To solve this, OSPF is uniquely designed to be Hierarchical. Network engineers divide the company into Areas.
- Area 0 (The Backbone): The core routers at Headquarters.
- Area 1: The New York Office routers.
- Area 2: The London Office routers.
*The Rule:* All traffic must pass through Area 0. If a cable breaks in Area 1 (New York), only the routers inside Area 1 recalculate their SPF algorithm. London (Area 2) doesn't care. They just know that Area 0 will eventually get the packet to New York. This hierarchical design allows OSPF to scale infinitely.
7. OSPF Example (Routing Table)
If you look at a table running OSPF, the Administrative Distance (AD) is 110.
text
-
O: Learned via OSPF.
-
[110/65]: Administrative Distance is 110. The calculated SPF Metric (Cost) is 65.
8. Best Practices
- Design Area 0 Carefully: The most critical rule of OSPF architecture is that *every* Area must physically connect directly to Area 0 (The Backbone). You cannot daisy-chain Area 2 to Area 1 to Area 0. If you do, OSPF will completely break and refuse to route the traffic.
9. Common Mistakes
- Mismatched Timers: OSPF routers form "Neighbor Adjacencies" by sending tiny Hello packets to each other. If Router A sends a Hello every 10 seconds, but a junior engineer configures Router B to expect a Hello every 30 seconds, they will refuse to become neighbors. They will ignore each other's LSAs, and the network will drop offline.
10. Mini Project: Calculate the SPF Cost
Imagine a triangle: R1, R2, R3.- R1 to R3 direct link: Very slow DSL (Cost = 500).
- R1 to R2 link: Fast Fiber (Cost = 10).
- R2 to R3 link: Fast Fiber (Cost = 10).
11. Practice Exercises
- 1. Explain why OSPF requires significantly more router CPU processing power and RAM than RIP.
- 2. What is the architectural purpose of restricting all OSPF inter-area traffic to flow exclusively through Area 0?
12. MCQs with Answers
Question 1
What is the specific metric utilized by the OSPF routing protocol to calculate the Shortest Path First (SPF) to a destination network?
Question 2
In a multi-area OSPF enterprise architecture, which area serves as the mandatory, central transit hub that all other areas must connect to?
13. Interview Questions
- Q: Contrast Link-State protocols with Distance Vector protocols regarding topological awareness and convergence time.
- Q: Explain the function of Link-State Advertisements (LSAs). How do they prevent the massive bandwidth overhead seen in older protocols like RIP?
- Q: Walk me through the mathematical logic of the SPF (Dijkstra's) algorithm when presented with a 1-hop low-bandwidth path versus a 3-hop high-bandwidth path.