[j-nsp] Next hop self for BGP

[Mark Tinka]

On Monday 26 April 2010 11:30:08 am David water wrote:

> I guess there are multiple way of changing next-hop
>  attribute to self. One write import policy with
>  next-hop-self and apply under EBGP group, other is write
>  export policy under IBGP group and other is at neighbor
>  level. Anything else? Can some one describe it when and
>  which one to use?

Having started with IOS (as I'm sure many of us on the list 
did), we tried to keep it simple by replicating IOS 
configurations when introducing JUNOS into the network.

As such, we've always modified the NEXT_HOP attribute on 
export from the PE router toward the route reflector. Other 
methods should work, but we haven't had a reason to try 
them.

I guess one issue I could see with applying this policy on 
eBGP ingress is if you have different import policies for 
each eBGP peer on the same router (which is not uncommon). 
This would mean each individual import policy for each eBGP 
peer would need to be configured to modify the NEXT_HOP 
attribute, while implementing this on export toward to the 
route reflector (iBGP) covers ALL eBGP peer NEXT_HOP 
information in a single configuration (of course, this could 
be simplified using groups or having a standard import 
policy for each eBGP peer, whichever is more feasible).

Again, we haven't had to do this so I'm just theorizing.

Cheers,

Mark.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: This is a digitally signed message part.
URL: <https://puck.nether.net/pipermail/juniper-nsp/attachments/20100426/7ccfb48d/attachment.bin>