OSPF routers must form an adjacency before they exchange routes, and the state a neighbor is stuck in tells you almost exactly what's wrong. The goal state is Full; anything else is a clue, not a mystery.
1. Read the neighbor state
show ip ospf neighbor # the state column is everything show ip ospf interface # area, timers, network type, MTU
Map the stuck state to its cause: Down/Init = hellos aren't being received back (one-way — ACL, interface, or auth); 2-Way stuck on a broadcast segment can be normal (DROTHER-to-DROTHER), but unexpected means a DR election issue; ExStart/Exchange = an MTU mismatch (the classic) or a duplicate router-id.
2. Check the must-matches
Two OSPF routers only become neighbors if a specific set of values agree on the link: area ID, hello/dead timers, authentication, subnet/mask, MTU, and network type. Walk that list against show ip ospf interface on both ends — a single mismatch blocks the adjacency.
3. Confirm reachability and IDs
Make sure hellos can actually flow — the interface is up, in the OSPF process, and not blocked by an ACL — and that no two routers share a router-id. Duplicate router-ids wedge the adjacency and log noisily.
4. Force a clean re-adjacency
Once the mismatch is fixed, restart the OSPF process (or bounce the interface) so the neighbors renegotiate from scratch rather than clinging to stale state, then watch show ip ospf neighbor walk up to Full.
Practise network incidents
OSPF adjacency loss is one of the router incidents in the simulator (needs the Network Diagnostics skill to reveal the hint). Work it in a full incident response shift, alongside BGP and STP storms.