> >
> > I'm from the camp of thinking that originating your routes
> > via BGP from your own AS is the cleanest way to multihome;
> > however, if you're just recieving some sort of default from
> > your upstreams, all above still holds.
>
> I am sorry, I thought we were discussing scenarios where multihoming with
> BGP was not possible due to hardware contraints.
Perhaps it is, but as I said, in the case were you've got
equal cost default routes (BGP-derived or static) the same
switching mechanism is used for outbound traffic and the
reasoning should be the same. Per-destination (or per-src/dest
pair in the case of CEF) load-sharing should yield better behavior
than per-packet as long as you don't overload one of the links
to the point of packet loss. If you do, try both ways to see what
works best for users and think about upgrading bandwidth.
Tony
>
> -%
> Cliff Judge Network Engineer
> 301-598-0500 x2866 Cidera, Inc
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:18 EDT