Week 1: OSPF

Contents
Week 1: OSPF..............................................................................................................................................1

OSPFv2.....................................................................................................................................................2

Gateway Protocols...............................................................................................................................2

Components of OSPFv2.......................................................................................................................2

OSPFv2 Message Format.....................................................................................................................3

Hello Packets........................................................................................................................................3

OSPF Designated Router(DR) and Backup Designated Router(BDR)....................................................4

Link State Operation............................................................................................................................4

OSPFv2
Open shortest path first is used for routing in IPv4 Networks, This was developed as an alternative to RIP.

Gateway Protocols
When travelling on a single autonomous systems we use Interior Gateway Protocols split into Classful
(RIP) and Classless(RIPv2, OSPFv2, EIGRP) IPv6 has its own protocols such as OSPFv3 or EIGRP for IPv6

- Classless – Supports VLSM and CIDR

- Efficient – No periodic updates, uses triggered updates such as a new device added or removed

- Fast Convergency – Quickly propagates network changes

- Scalable – Works well in small and large networks, Routers can be grouped into Areas and
support hierarchical topology

- Secure – Supports Message Digest 5 (MD5) authentication for routing – Stops man in middle
attacks as routers must prove their allowed on network

Components of OSPFv2
Databases

- Adjacency Database – show ip ospf neighbor

- Link-state Database – Show ip ospf database

- Routing Information Database – show ip route

Message Types

- Hello – Looks for neighbours to exchange information

- Database Description – Describes network to each other such as other routers

- Link-State Request

- Link-State Update

- Link-State Acknowledgement

Algorithm

- Dijkstra’s SPF Algorithm – Trys to find shortest path by cost

OSPFv2 Message Format

LSU or link state updates will be sent to send link changes on the network

Hello Packets
Hello packets are used for many things with a 32 bit packet containing Router ID and Area ID

Network mask, interval for sending another hello, Dead interval to wait if you don’t receive a hello from
the neighbour to say its down, Designated routers (DR) and Backup Router (BDR) and list of neigbours

To establish a Adjaceny in OSPF All items in the hello packet must match, such as Dead interval and
Subnet

Hello packets are used for many things

- Discover OFSP Neighbours

- Establish Neighbourhood Adjacency

- Elected Designated Router(DR) and Backup(BDR)

OSPF Designated Router(DR) and Backup Designated
Router(BDR)
Multiaccess networks create two issues for OSPF and the flooding of LSA packets

- Creation of Multiple Adjacencies – Leads to excessive number of LSA packets being exchanged

- Extensive Flooding of LSA – When OSPF is initilaised link-state routers will flood the network

The solution to this is

- Assigning a Designated Router and Backup

o This makes it so routers only form adjacency to the two designated routers

- All other routers become DROTHER and only form adjacency with DR and BDR

- OSPF DR is the collection and distribution of point for all LSA packets sent to stop flooding

o All LAS are sent on 224.0.0.6

o DROTHER receive via 224.0.0.5

- BDR automatically assumes the DR role if the DR fails

Link State Operation

- Establish Neighbour Adjacency

- Exchange Link-State Advertisements

- Build the Link State Database

- Execute the SPF Algorithm – Shortest Path First

- Choose Best Path – lowest cost based off bandwidth, etc

Operation State
- Down State – No hello sent, Transition into init

- Init state – Hello packets sent

- Two-Way State – Communication is bidirectional, On multicast networks routers elect DR and

BDR

- ExStart State – Initiates DBD packet exchange

- Exchange State – Exchanging DBD Packets

- Loading State – LSR and LSUs are used to gain additional route information, Routers are using
SPF Algorithm

- Full State – Databases are fully synchronized

Establish Adjacencies
- Loopback ID is used to establish Router ID

- If there is no loopback use the highest active ipv4 address

- Higher Router ID is priority and will start exchange

Synchronizing OSPF Database

After the two wat state routers transition to database sync this is a three step process

- Highest Router ID sends DBD First

- Exchange as many DBD as needed to convey database, Router must acknowledge each DBD with
an LSAck Packet

- Exchange LSR between routers compressing DBD with local LSDB, If the DBD has more current
link information the routers transition to loading state then moving to full state

After all Link State Requests have been fulfilled updates will only be sent when

- A change in the network is perceived such as new device

- Every 30 Minutes
Single area and Multi area OSPF
Single Area – All Routers are in same area, Best Practice is to use area 0

Multi Area – All areas must connte t to backbone area (0) Routers interconntecitng Areas is known as
Area Border Routers ABRs

Multi Areas have many advantages

- Smaller Routing Tables

- Reduced Link-state update overhead

- Reduced frequency of SPF Calculations