El 16/06/2006, a las 23:31, Iljitsch van Beijnum escribió:
On 16-jun-2006, at 20:03, marcelo bagnulo braun wrote:i guess it is mostly a matter of how to define a multihomed hostwe could say that we have two types of multihomed hosts:- the "strict" multihomed host where the host has multiple physical interfaces- the "loose" multihomed host, where the host has multiple addressesIn the first case it is pretty clear that the host multiple paths to connect to the internet, since the attachment point is different.Actually, no. You assume that different interfaces will lead to different off-site paths.
not really, i am assuming that a host with multiple interfaces has multiple paths to connect to the Internet, since the host itself has as many attachment points as it has interfaces
But it's also very possible that different physical interfaces are just used for redundancy towards the rest of the site,
which is part of the rest of the internet from the host's perspective, so a host with multiple interfaces has multiple paths to the rest of the internet (that is what i meant..)
anyway, i guess we all understand what we are talking about here, it is just a matter of terminology
what it is also clear i guess is that terminology we are using is not very precise... :-(
regards, marcelo
regardless of whether the site is multihomed or not. This is what I consider to be "host multihoming" although I think the current discussion shows that using this terminology isn't too fruitful. This type of host multihoming has no impact on global routing because the different addresses for the host are aggregated into a single prefix belonging to the site. (Or this all happens at the ethernet link level and there's only one IP address. With FDDI there was multihoming at the physical level, very cool.)