OSPF overview

OSPF (Open Shortest Path First) is the link states routing protocols. Because it’s an open standard, it’s implemented by variety of the networks vendors. OSPF will be run on the most routers that does not necessarily have to Cisco routers (unlike EIGRP which can run only on the Cisco routers).

Here is most important feature of the OSPF:

·         A classless routings protocols
·         Support VLSM, CIDR, manual route summarizations, equal cost loads balancing
·         Incremental update is Supported
·         Uses only one parameters as metric
·         The administrative distances of the OSPF routes by default, 110.
·         Uses multicasts addresses 224.0.0.5 and 224.0.0.6 for routings update.

Router running OSPF have to the established neighbour relationships before exchanging routes. Because OSPF is a link state routings protocols, neighbours do not exchange routings tables. Instead, they are exchange information’s about the networks topology. Each OSFP router then runs SFP algorithms to the calculate the best routes and adds those to routing tables. Because each routers knows entire topology of the network, the chance for routing loop to the occur is minimal.

Each OSPF routers stores routing and topology information’s in the three tables:

·         Neighbour table – The stores information’s about OSPF neighbours

·         Topology table – The stores topology structures of the network

·         Routing table –  The stores best routes

OSPF neighbours

OSPF routers need to the establish neighbour relationships before the exchanging routing updates. OSPF neighbours are the dynamically discovered by the sending Hello packets out each OSPF-enabled interfaces on the router. Hello packets are sent to the multicast IP address of the 224.0.0.5.

The process is explained in following figures:

Router R1 and R2 are directly connected to each other’s. After OSFP are enabled both router send Hellos to each other’s the establish a neighbour relationships. You can verify that neighbour relationships has indeed been the established by typing the show ip ospf neighbours commands.

In this examples above, you can see that router-id of R2 are 2.2.2.2. Each OSPF router assigned a router ID. A router ID is the determined by using one of following:

1.    Using the router-id commands of the OSPF processes.
2.    Using the highest IP address of router’s loopback interface.
3.    Using the highest IP address of router’s physical interface.

The following fields in Hello packets must the same on both routers in the order for routers to the become neighbours:

·         Subnets

·         Area id

·         Hello & dead interval timers

·         Authentications

·         Area stub flags

·         MTU

 

By default, OSPF sends hello packets every 10 second on the Ethernet networks (Hello interval). A dead timer is the four times value of hello interval, so if router on an Ethernet network does not receive at least one Hello packet from OSFP neighbour for 40 seconds routers declares that the neighbour to be down.

OSPF neighbour states

Before establishing the neighbour relationships, OSPF routers need to the go through several state changes. These states are explained below.

1. In-it state – A router has received Hello message from other OSFP routers
2. 2-way state – The neighbour has received the Hello messages & replied with Hello message of his own
3. Ex start state – Beginning of LSDB exchanges between both routers. Router are starting to the exchange link state information’s
4. Loading state – One neighbour sends LSRs (Link State Requests) for every networks it does not know about. The other neighbour replies with LSUs (Link State Updates) which the contain information about requested networks. After all requested information have been received, other neighbour goes to the through the same processes
5. Full state – Both router have the synchronized databases and fully adjacent with each other’s.