OSPF neighbor relationship process

1. Determine the Router ID

It is the router’s name in the OSPF process. It’s always advisable to hard code the router-id.
Router-id > Loopbacks > Active Physical interface IP

2. Add interfaces to the link state database, done by the network command

3. Send a Hello message on chosen interfaces

Broadcast/P-2-P networks	Every 10 seconds
Non-Broadcast Multi-access Networks (NBMA)	Every 30 seconds

Router ID	Either hard coded or interface IP address
Hello and Dead timer	Based on whether it’s a Broadcast/P2P/NBMA network
Neighbors	If the neighbor is present, it knows it is not a new relationship and resets the dead timer. If not present then it adds the neighbor.
Router Priority	Determines the Master/Slave router on the network segment
DR/BDR IP address
Authentication password	The authentication password should be the same, obviously.
Network Mask	On both the neighboring interfaces should be the same
Area ID	Area of both the to-be-neighbors on that segment should be the same

Broadcast/P-2-P networks	Every 40 seconds
Non-Broadcast Multi-access Networks (NBMA)	Every 120 seconds

4. Receive Hello (INIT state)

If the state flaps between INIT/DOWN check the things that need to match on both the neighbors.

5. Send Hello Reply (2-WAY state)

Checks the hello packet received from the neighbor to see if it is listed as a neighbor or not;
- If yes, resets the dead timer and continues to be an existing neighbor.
- If no, adds as a new neighbor and proceeds further to other checks.

6. Master/Slave relationship determined (EXSTART state)

7. DBDs are acknowledged and reviewed (LOADING state)

While reviewing the DBDs if the Slave realizes that there is information missing for say 2 subnets/prefixes, the Slave will send 2 Link State Requests (LSR) and Master sends back 2 Link State Updates (LSU).
If Master router realizes that it is missing some info about a particular prefix, it will query the slave with an LSR and the slave with reply back with an LSU containing the LSA for that prefix.

8. Neighbors are synchronized (FULL state)

Now the SPF algorithm can be run on the LSDB that the routers now have, and they can then figure out the best path to each prefixes and put them in the routing table.

Networkology!