[c-nsp] Loopback Advertise in OSPF

[Phil Mayers]

Mark Tinka wrote:
> On Wednesday 27 February 2008, Phil Mayers wrote:
> 
>> I think that's probably a bit dramatic (no offence
>> intended). We've used "redis connected / static" for
>> "customer" routes (edge networks) for quite a while now
>> here with no problesm; initially in a VRF-lite model:
> 
> If you read further on in my post, you will see I said 
> redistribution is bad with the exception of MPLS/VPN's 
> and/or VRF's.

Yes I read that, but I am having a hard time seeing what the fundamental 
difference is between "redis connected" inside a VRF versus not (for the 
same protocol). Surely if one is bad, the other is?

> 
>> * "redistribution is bad" - which is certainly true 
>> *between* the  
>> dynamic protocols, but no-one has ever convinced me that's 
>> true for 
>> connected or static routes
> 
> Well, I wouldn't be keen on having flapping customer links 
> keeping my OSPF process busy. Keeping OSPF light by using 
> it for Loopback and core interfaces only is how I'd deploy 
> it.

Sure. No argument there - that is why we made the move.

> 
> Using iBGP to originate customer routes means you don't need 
> to redistribute any connected prefixes into routing 
> protocols.

So this is ok:

router bgp 6xxxx
  address-family ipv4 vrf CUST1
   redis connected

but this isn't:

router bgp 6xxxx
  address-family ipv4
   redis connected

Why?

I'm not disputing you - I'm genuinely curious why one way of doing 
things is better.