North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: Why Vadim likes statics

  • From: Robert E. Seastrom
  • Date: Mon Apr 24 18:00:30 1995

> This sounds a lot like the slippery slope of static routing being the 
> most stable, so we should encourage its use Internet wide.   I -know- 
> Karl D. 
> (and others that depend on dynamic routing for alternate provider 
> fallback) will kick at this. 

On the contrary, Bill, being set up to propagate IGP information to your EGP 
in situations where not absolutely necessary seems to be a perversion of the 
whole intent of EGPs vs. IGPs.  Perhaps it's time to rethink for a moment why 
the whole Internet isn't just running one big IGRP or OSPF cloud.

Kudos to the people who have realized that for singly-connected networks, the 
BGP advert. status of the network should change precisely as often as the 
policy (eg. ONCE, when you set the network up, and AGAIN if they change to 
another provider).  Granted, there are a lot of people who are _not_ singly 
connected, but they (folks like Karl D.) are (a) running BGP themselves, and 
(b) sophisticated enough to pull up their own networks at the periphery of 
their cloud.

The point I'm making is that you shouldn't be BGP peering with customers who 
are not multiply attached (that's what IGPs are for), and you should not be 
propagating your IGP information to your BGP except in rare cases.  Let IGPs 
do what they are best at (dynamic information and reachability on a micro 
level) and EGPs do what they're best at (reachability and policy on the macro 
level)  Simple, eh?

By the way, yes, I am biased.  I have been saying that this is the only sane 
way to do routing adverts to other providers for neigh unto two years now.


Robert E. Seastrom -- [email protected]
Network Engineer, Digex International
My posting, my opinions, not speaking for the company, etc. etc.