Stefano, all, On Wed, Feb 01, 2012 at 06:04:23PM +0100, stefano previdi wrote: > > Btw: which application needs to know the topology? > > E.g.: a CDN may want to operate autonomously with maps rather than > using ECS.
Downloading a map and operating autonomously with it is one thing, trying to derive missing fields in the map/matrix by some kind of triangulation is another thing. > Not that I'm advocating this but it all depends on where > do you want to place the complexity of topology hint: inside ALTO > server (ECS like of approach) or inside the application (Maps that > allows you to compute paths/trees). Maybe part of the confusion is that we need a clear definition what end-to-end is. I'd say that ALTO should always give the end-to-end costs for sending IP packets between two endpoints using "normal" IP-layer forwarding. Based on that knowledge one can of course build efficient paths/trees that involve proxies, media relays or the like (i.e., this higher-layer path is a concatenation of several IP-layer end-to-end paths). Does this make sense? > > If I remember > > correctly once upon a time our primary goal was assisting peer selection > > (i.e., I know that I can get the desired resource form IP address A, B, > > or C, so please ALTO service tell me which address I should give a > > connect()-try first). To that end, the 1xN vector would be enough. > > The CDN is not substantially different in its logic. this is my feeling as well, that's why I was asking. Thanks, Sebastian _______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
