[c-nsp] Loopback Advertise in OSPF

[Darryl Dunkin]

I've always avoided that. Wouldn't it attempt to run the OSPF process on
the loopback interface? There aren't any adjacencies possible there so
it is a waste of resources I would think (however minimal it may be).
Setting it passive would avoid this of course, but is a pain to manage
for large amounts of interfaces (ok ok, Cisco does have
'passive-interface default' now). That depends on if you are using your
IGP only for BGP peering via loopbacks, I need all connected interfaces
advertised into my IGP (non-core, customer, loopback), without running
the process across them, and without wildcarding 0.0.0.0 for all
interfaces (diverse amount of subnet numbering here).

I'd like to be corrected if this isn't the way to do it for any reason
though.

-----Original Message-----
From: Phil Mayers [mailto:p.mayers at imperial.ac.uk] 
Sent: Tuesday, February 26, 2008 15:03
To: Darryl Dunkin
Cc: Rupert Finnigan; cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] Loopback Advertise in OSPF

Darryl Dunkin wrote:
> If you're using /32 masks for your loopbacks (as you should):
> router ospf ####
>   redistribute connected subnets 
> 
> The key part is to define 'subnets'.

My personal preference has been to allocate router loopbacks *and* p2p 
IPs out of a cidr block and use a network statement (which can be 
identical on all routers):

int lo1
  ip address 192.168.0.1 255.255.255.255
int vl4000
  ip ospf network point-to-point
  ip address 192.168.1.1 255.255.255.252

router ospf 1
  network 192.168.0.0 0.0.1.255 area 0