OSPF Neighbors and Adjacencies

Three neighbors, all in FULL state — the goal. Two are on a broadcast network (Gi0/0) with DR and BDR roles; one is on a point-to-point link (Gi0/1) with no DR role. FULL means the LSDBs are synchronized between the routers.

#The OSPF State Machine#link

The state progression is deterministic. Down means no Hello received. Init means we got a Hello but it didn't include our own Router ID — the peer hasn't heard us yet. 2-Way means bidirectional communication is confirmed. ExStart through Loading is database exchange. Full is when they're synchronized.

#The Hello Protocol#link

• ▪Sent every 10 seconds on broadcast and point-to-point links (30 on NBMA)
• ▪Contains: Router ID, Area ID, Hello/Dead intervals, network mask, priority, DR/BDR
• ▪Dead timer is 4x Hello by default — 40 seconds
• ▪If Hello not received within Dead interval, neighbor is declared down

Every Hello carries parameters that must match between the two routers. If Area ID, Hello interval, Dead interval, network mask (on broadcast), or authentication doesn't match — the adjacency never forms. This is the #1 cause of neighbor failures, and it's always a config mismatch.

#The Matching Requirements#link

#Diagnosing the Failure#link

Stuck in ExStart/Exchange almost always means MTU mismatch. The two routers can't agree on who's master and can't exchange database descriptors. Check `show interfaces` on both ends for MTU — one might be 1500 and the other 1400 due to a tunnel or jumbo frame config.

#Check Your Understanding#link

#Common Mistakes#link

• ▪Mismatched Hello/Dead timers — the most common config error
• ▪Area ID typo — one side says area 0, the other says area 1
• ▪MTU mismatch — usually from tunnel interfaces or differing MTU configs
• ▪Authentication mismatch — key or type differs
• ▪Assuming FULL is required on broadcast networks — 2-Way with non-DR/BDR is healthy