OSPF Troubleshooting

[TRACY HARTMANN]

We have a small setup for OSPF between two internet routers and a Firewall.  We started having some drop issues today and I saw the dead times keep changing, not sure if that is a sign of an issue or just a common ocurrance.

router#show ip ospf neighbor

Neighbor ID Pri State Dead Time Address Interface
X.X.X.x  1 FULL/BDR 00:00:39 C.C.C.C  TenGigabitEthernet0/0/0
X.X.X.255 FULL/DROTHER 00:00:32 c.c.c.255 TenGigabitEthernet0/0/0

then I show again and the timers change to less and they never get more than what is above.  The State never changes from Full and I tried a couple debugs and show nothing.   Just looking for some direction of the dead timers changing is common or if they should be staying stable?

[Deepak Kumar]

Hi,

I hope everything is fine in the OSPF. As you shared some details which are looking very important as

Neighbor ID Pri State Dead Time Address Interface
X.X.X.x  1 FULL/BDR 00:00:39 C.C.C.C  TenGigabitEthernet0/0/0
X.X.X.255 FULL/DROTHER 00:00:32 c.c.c.255 TenGigabitEthernet0/0/0

By default, the dead timer interval is four times the hello timer interval means 40 seconds. Again, after 10 seconds it will receive the hello and dead timer reset to 40 seconds. I can say that in a health lab or network dead timer will not go less than 30 seconds.

Now, As I am looking that Status of peer also good because you lab might be like:

RouterA is connected to a Switch and Router B and Router C is also connected to the switch and all are in the same network.

If yes then, here is the meaning of by you can not see Full/DR on this router because this router is elected DR in the network and another two routers elected as Backup DR (X.X.X.x ) and  FULL/DROTHER (X.X.X.255) and it is default behavior. 

FULL/BDR: Backup Designated Router (BDR) is elected to replace DR should DR fail. Hence DRother routers also form an adjacency with BDR.

FULL/DROTHER: DRothers will form an adjacency with DR and BDR only. DRother is a router that’s neither a DR nor a BDR.

As you are facing drops then please share logs and OSPF outputs. Have you noticed any Drop on the interface?

[balaji.bandi]

Those times are correct, OSPF has the ability to check neighbour live dead timers, they count down and keep checking hello-interval that is normal.

Since you configured broadcast domain not as a point to point they elect DR / BDR process to become DR this is normal in OSPF process.

if you looking to FULL/ FULL - you need to configure the interface as point-to-point

Drops could be the reason lost connection, what is your Logs show, is the OSPF neighbourship DOWN and UP ?