[c-nsp] OSPF NSSA Question

[Dan Armstrong]

In answer to the first question, I could use different subnets - but 
that seems like a huge waste of address space....

I would go through twice the IPs...




Niels Bakker wrote:

>>>My problem is that when I do that, the two distribution layer routers
>>>form an adjacency with each other over the link%$#@#!!
>>>
>>>I realize that is "normal" behaviour... but it is not deseriable..
>>>      
>>>
>
>How would packets flow between these routers if they were to use
>IP addresses from different subnets?
>
>
>  
>
>>>Can I:
>>>
>>>-Explicitly force the two dist routers somehow not to form the
>>>adjacency?
>>>      
>>>
>
>* markom at pangalactic.net (Marko Milivojevic) [Sat 11 Sep 2004, 00:11 CEST]:
>  
>
>>   This is interesting question, indeed. The only thing that comes to my 
>>mind is inbound ACL on one (or both) of the distribution routers blocking 
>>OSPF traffic from another distribution router...
>>
>>   Of course, there might be more elegant solution (this one strikes me as 
>>ugly, frankly).
>>    
>>
>
>Sure.  Declare the interface as passive and explicitly list your
>neighbors by IP address:
>
>!
>router ospf 1
> passive-interface FastEthernet0/0
> neighbor 1.2.3.4
> neighbor 1.2.3.6
>!
>
>
>	-- Niels.
>
>  
>