Re: Transport level multihoming

[Peter Tattam <peter@jazz-1.trumpet.com.au>]

On Tue, 7 Aug 2001, Ramakrishna Gummadi wrote:

> 
> 
> On Wed, 8 Aug 2001, Peter Tattam wrote:
> 
> > On Tue, 7 Aug 2001, Ramakrishna Gummadi wrote:
> >
> > >
> > > Another problem with the transport-level multihoming is that
> > > the number of failure modes in multi-tier multihoming are decreased. For
> > > example, consider the scenario where a site is dually-multihomed, while
> > > each of the providers are triply multihomed to three different providers
> > > each. The question how many addresses do the hosts in the site have? If it
> > > is two, then it means that the site can not tolerate failures in four of
> > > the providers twice removed from it. If it is six, then it means we are
> > > exposing the site to addressing issues not related to it.
> > >
> > > A tunneling+translation solution solves this problem cleanly by using only
> > > two (or
> > > as many addresses as there are providers) provider-independent addresses
> > > to the site. The site's immediate providers would independently translate
> > > the addresses to meet their own load-balancing needs without leaking the
> > > site's prefixes. When a failure between the provider and its higher-up
> > > is detected, by creating two tunnels between itself and the higher up via
> > > the two working ISPs, a provider can ensure connection survivability and
> > > can even load-balance failed traffic among the working providers. The
> > > tunnels are only one "leve" deep, and are transparent to all others,
> > > including the original site.
> > >
> > >
> > > thanks,
> > > ramki
> > >
> >
> > Won't a translation solution have the risk of translator box meltdown?  i.e.
> > may not scale?
> 
> It is only for the customer's prefixes that a provider has to make
> translation. The translation can reside in the customer's border router,
> and needs to contain only entry per site. Secondly, translation is
> entirely optional. Finally, I think stateless (with regard to
> packets) translation is not very expensive---already, web switches provide
> hardware layer-7 (looking at cookies and urls) "virtual ip" services for
> load-balancing and failover, so a layer 3 translation certainly looks
> certainaly possible to me.
> 

But it's still another single point of failure.  In my experience, such a
service unless delivered in a router style high reliability platform is going
to be a high risk point of failure.  Also the suggestion that it be another box
separate from the router means that it adds significant cost to the provider's
infrastructure.  It will be a specialized piece of equipment and likely to be
expensive.

Routers are good - they have high utility/cost, manage redundacy well, are
general purpose and you normally have more than one of them.

Peter

--
Peter R. Tattam                            peter@trumpet.com
Managing Director,    Trumpet Software International Pty Ltd
Hobart, Australia,  Ph. +61-3-6245-0220,  Fax +61-3-62450210