IN OSPF, when the designated router (DR) is lost for a particular segment, the backup designated router (BDR) then becomes the DR and a new BDR is elected. If the original DR then comes back online, there will be NO election, as the new DR and BDR are in place and functioning. The only way around this is to make sure that the hub site routers can only become DR and BDR for these segments. To make sure this is the case you have to set the OSPF priority to zero on all the interfaces at the spoke routers. This command effectively tells OSPF that this interface cannot become a DR or BDR for this segment. The command to do this is below and is placed on the appropriate interface:
Router(config-if)#ip ospf priority 0
After this change has been made to all the affected routers, you will have to issue the command 'clear ip ospf process' and restart OSPF. If this command is not available in the IOS image you are running, removing and readding the OSPF config will fix it, otherwise a reload is in order and you must write the configuration for the changes to take affect.
Remember that you can set the OSPF priority on the hub routers as well - but the priority HAS to be between 1 and 255. The higher the priority, the more likely that the router will become the DR. OSPF network-types that require a DR/BDR have their interface priority set at 1 by default, and after the election, the DR's priority jumps to 200.
