Multihomed to 2 ISPs - Load Balance?

Mon Jun 26 07:28:02 UTC 2006

On Mon, 26 Jun 2006 06:28:41 +0000 (GMT), "John Smith"
<jsmith4112003 at yahoo.co.uk> said:
[snip]
> 
> > There isn't a facility in bgp to tell a neighbor more than one possible
> > aspath... or not one that most network folk use currently.
> > 
> > I suppose for a subset of routes you might hack up some community based
> > solution, but it'd be a horrible hack, and it'd cause you to keep churning
> > your router configs on a very regular basis as things up stream changed.
> > 
> > If the downstream has a connection only to you does it matter where they
> > send packets? everything has to go through your AS to get anywhere...
> > right? If they have a multihomed solution (you and another isp) they are
> > going to have to decide on some other internal metric (interal to them
> > based perhaps on non-routing-table information, like 'john has a oc-12 to
> > provider-Y, Jim only has a T1.... send to John!') whete to send traffic.
>  
> I dont think its as simple as this. In the simplest case, assume that my
> downstream peer has a policy to reject all routes that traverse AS 20.
> Now i am splitting all load across AS 10 and AS 20, while i tell him that
> i'll be only sending the traffic through AS 10. This can create problems
> for my downstream peers, and in the worst case can lead to
> blackholes/loops.

If an AS is balancing traffic outbound in weird ways that's only a
problem for their customers if it doesn't work. Customers do otherwise
not see any details beyond the ability to filter based on AS-paths they
receive, so outbound there's no potential for "blackholes" with current
standards which specify that only the best (one) path is announced. 
For inbound traffic you'd only have a problem if every packet did 
carry a recorded traversal path and there were equipment around that 
could make decisions based on such information. 

What you're asking for simply doesn't exist in today's network. 

Those who want to the ability control path-selection globally
should participate in IETF workshops to get such functionality
included in future network-architectures ;)

//per

-- 
  Per Heldal
  http://heldal.eml.cc/