Routing

Static routing

Dynamic routing

• Minimum configuration and setup required
• Routers are automagically discovered and maintained through the exchange of routing information
• Routing decisions are made on dynamically learned routes

There are two classes of routing protocols --- distance vector and link state.

Distance vector routing

• Each router informs neighbors of its entire routing table periodically
• Routing table consists of
  • Destination
  • Cost
• Router selects neighbor advertising lowest cost

Once all routers have a full view of the network, the network is said to have converged

Routing Information Protocol (RIP)

• Uses hop count as cost
• Max hop count is 16

Routing table

Network	Destination network address	Next hop IP / outgoing interface	Metric
A	192.168.5.0	Directly connected	2

Open Shortest Path First (OSPF)

• Uses cost as metric
  • Speed
  • Delay
  • Reliability

Comparison with RIP

• More realistic factors taken into account to calculate cost
• Updates as soon as there's a change instead of at intervals
  • Faster convergence
• No hop count limit
  • Suitable for large networks
• Less bandwidth consumption

Comparison

	Static routing	Dynamic routing
Advantages	Traffic can be forced to take a more secure path Less network overhead (no need to send routing information)	Lesser configuration compared to static routing Able to configure redundant paths with STP/RSTP
Disadvantages	Each network has to be configured manually -> less flexibility No redundancy for failure of links Need to reconfigure most routers if there's a change in network Not suitable for large networks	More overhead traffic (need to send routing information) RIP: every 30 seconds Unable to manually specify routes -> less flexibility compared to static routing

---

Last update: June 11, 2023
Created: June 11, 2023