OSPF some prefixes in DR routing table have more that 1 source.

[elawton1986]

I've been browsing for information on this issue and do not see where it is explained.

Any thing revelant is greatly appreciated.

The DR has many neighbors as remote locations.  Most neighbors have 1 to 2 local networks.  There are 2 ipsec/GRE tunnels for each remote neighbor

connecting to the DR and BDR.  The local networks at the remote locations are unique. Some of these networks show as having mutiple advertisments in the DR. here is an example:

sh ip route

       10.25.6.0/26 [110/2010] via 10.25.6.89, 1d01h, Tunnel8

                          [110/2010] via 10.25.6.34, 11:03:53, Tunnel11

                          [110/2010] via 10.25.6.81, 1d01h, Tunnel7

                          [110/2010] via 10.25.6.7, 1d01h, Tunnel9

sh ip route 10.25.6.0

Routing entry for 10.25.6.0/26

  Known via "ospf 1", distance 110, metric 2020, type intra area

  Last update from 10.25.6.30 on Tunnel7, 1d01h ago

  Routing Descriptor Blocks:

    10.25.6.89, from 10.25.6.1, 1d01h ago, via Tunnel8

      Route metric is 2020, traffic share count is 1

    10.25.6.81, from 10.25.6.1, 1d01h ago, via Tunnel7

      Route metric is 2020, traffic share count is 1

  * 10.25.6.34, from 10.25.6.1, 1d01h ago, via Tunnel11

      Route metric is 2020, traffic share count is 1

    10.25.6.7, from 10.25.6.1, 1d01h ago, via Tunnel9

      Route metric is 2020, traffic share count is 1

The remote network 10.25.6.0/26 is at the remote site connected via Tunnel11

Thanks for looking into this.

[Richard Burts]

We do not have information about the topology of the network and this prevents us from having a complete understanding of the issue. But clearly what is going on is that 4 routers are reporting that they have a route to network 10.25.6.0/26 and all 4 routes come into the DR router with the same metric (2010). So OSPF puts all 4 entries into the routing table since they all have equal metrics. This results in load sharing over the 4 connections.

Perhaps if you supply more information about the topology of the network, configurations of the affected routers we might get a better understanding and be able to give you a better answer.

HTH

Rick