OSPF database exchange to DR and BDR

[Cisco Freak]

Hello Experts,

I have a query related to how the DB exchange will happen a broadcast OSPF network with 3 or more neighbors. I know that the routers will use a multicast address(224.0.0.6) send DBD to DR and BDR. And then DR will flood the LS data to 224.0.0.5.

But how does DR and BDR exchange the DB each other?

I took a Wireshark capture in GNS3 and that confused me. I see DBD going outbound from DROther device to IP address of the BDR and vice versa.

SW4#sh ip os neighbor 
Neighbor ID     VRF      Pri State                  Dead Time   Address         Interface
10.1.1.3        default  0   FULL/DROTHER           00:00:32    10.1.1.3        Ethernet1
1.1.1.1         default  1   FULL/BDR               00:00:31    10.1.1.1        Ethernet1
2.2.2.2         default  1   FULL/DROTHER           00:00:34    10.1.1.2        Ethernet1
SW4#

Can anyone please explain how this process works? Link to any useful document is also appreciated.

[balaji.bandi]

From my notes (not like to re-invent the wheel)

The DR and BDR form a full adjacency with each other. Every other node forms a full adjacency with the DR and the BDR. DROTHERs only go to the 2WAY state with each other.

If an LSA is ORIGINATED on the DR (because the change happened on that router, or on something connected behind that router in a different subnet), the DR sends the update to 224.0.0.5 directly. All other devices respond with an acknowledgment, the DROTHERs do this on 224.0.0.6, the BDR on 224.0.0.5.

If an LSA is originated on the BDR, the BDR sends the update to 224.0.0.5 directly. All other devices respond with an acknowledgment, the DROTHERs do this on 224.0.0.6, the DR on 224.0.0.5.

If an LSA is originated on a DROTHER, the DROTHER sends the update to 224.0.0.6, and the DR sends it to all other routers, including the BDR, on 224.0.0.5. The BDR sends an ack, acknowledging the DR's rebroadcast on 224.0.0.5, the other DROTHERS send an ack, acknowledging the DR's rebroadcast on 224.0.0.6.