chapter 5 - network layer: control plane

Question 1

What is the goal of a routing protocol?

Answer 1

Find good paths from source to destination (least cost, fastest)

Question 2

How does traditional per-router control work?

Answer 2

Each router runs its own routing algorithm and interacts locally

Question 3

How is an internet control message protocol (ICMP) message delivered?

Answer 3

It is carried inside an IP packet
back to the original sender with error message

Question 3

How does logically centralized control (SDN) work

Answer 3

A central controller computes and installs forwarding rules in routers

Question 3

What info does an ICMP error message include?

Answer 3

shows the path your data takes from your device to a destination
—-
* Type, code, first 8 bytes of the IP datagram that caused the error

Question 4

How does traceroute use ICMP?

Answer 4

TTL=1, for each hop TTL=-1. when TTL = 0 drops packet and sends back an ICMP msg to sender, so sender knows which is the first packet. then TTL=2, hopper 2 hopp og sender ICMP message to sender which is router nr 2

(time to live) = TTL

Question 5

When does traceroute stop?

Answer 5

When the destination replies with “port unreachable” (ICMP type 3, code 3)

Question 6

What is network management?

Answer 6

Monitoring, configuring, and controlling network devices to ensure performance

Question 7

Difference between global and decentralized routing?

Answer 7

Global: all routers know the network topology and link costs
Decentralized: routers know only neighbors and iteratively update info

Question 8

What kind of algorithm is Dijkstra’s? + goal of the algorithm

Answer 8

Global link-state routing algorithm
The least-cost path from a source node to all other nodes

Question 9

Key difference between link-state and distance-vector routing?

Answer 9

Link-state: full topology knowledge, fast convergence
Distance-vector: local knowledge, simpler, but slower convergence