OSPF Hello behavior 2WAY to INIT, 1-Way

[Gunter Vlemmix]

I have noticed in a router log the following message:

%OSPF-5-ADJCHG: Process 100, Nbr 192.168.254.40 on FastEthernet0/0 from 2WAY to INIT, 1-Way

My question is now, when does the router goes from 2WAY to INIT, 1-Way state ?

The previous state was 2-Way.

INIT state means that I have received a hello packet but my routerId is not included in the received hello packet.

This means that I'm recieving hello packets from the adjacent router.

If the adjacent router doesn't see my hello packet, it should wait deadrouterinterval before tearing down the adjacency.

I don't expect the neigboring router that "misses" a hello packet from my router, to remove my RouterID in the hello packet.

So, I'm just asking myself under which circumstances we can receive this error.

Any idea's?

[smalkeric]

There are a few circumstances that may cause this 2 way to init state transition. 1.For any of the 2 routers to be neighbors, they should belong to the same area.Their router interval and dead interval should match. I think if the neighboring router is changed its area and at the same time, debug ospf adj is on.So that the 2- way state gets degraded to init state.There may also be some other reason.

Are you checking this in a lab environment with both the routers under your control?

[Gunter Vlemmix]

It is a live customer environment, were I have a mix of third party L3 switches and Cisco routers.

I have noticed that the problem only comes from the third party switches (Cisco just logs the error).

However, it's an operational environment, so all OSPF parameters are correct and don't change.

I'm just asking myself what a possible reason can be why my routerId is missing in a received hello packet while the adjaceny is up and running???