Sebastien Roy writes: > I don't see how this would directly lead to the conclusion that the IP > tunneling service needs to depend on milestone/network. Why wouldn't an > administrator view milestone/network as including the configuration of > IP tunnel interfaces?
Agreed; it should. Assuming there is a meaningful "milestone/network." (I'm not too sure there is, but it seems others are more sure.) > The IP tunneling service could very well depend > on an individual service that is responsible for "non-IP-tunnel" IP > interfaces rather than the broader milestone/network. That's the part I'd very much rather not have. I agree that we need to have some way of telling the user that his chosen tsrc and/or tdst are unusable at some point in time, and that therefore the tunnel itself is unusable. That does _not_ mean that the tunnel is permanently unusable (if tsrc is bad, we may acquire an IP address later that makes it good; if tdst is unreachable, we may acquire a route later that makes it reachable), nor does it mean that we ought to be representing it as a configuration failure. It's no more a configuration failure than is setting an IP address on an interface that currently lacks the right cable. The existence of the wire is an outside condition, not an internal configuration issue. Adding these sorts of internal dependencies just makes supporting normal networking protocols, such DHCP and routing, very much harder. Maybe just impossible. -- James Carlson, KISS Network <james.d.carlson at sun.com> Sun Microsystems / 1 Network Drive 71.232W Vox +1 781 442 2084 MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677
