On 30/07/13 10:53 -0400, Russ White wrote: > > > So the only way is to add a route pointing to an interface. This is very > similar > > to how static routes are configured on routers today. AFAIK, a lot of > > operators use static routes and if we get rid of the "interface" > > object, then a whole class of static routes cannot be programmed using the > > RIB IM model. Similarly, you won't be able perform things like RPF > > checkŠwithout which multicast will be crippled. > > I assume you're talking about a static to an interface. For this specific > case, I would argue that we should wait and see if that's really a > requirement --statics to interface aren't (normally) in operational best > practices, so I don't know the value of supporting them. For > point-to-points, just use the other end of the link, and for broadcast, a > static to an interface is generally not a really good idea (ARP cache).
I'm not sure I agree with this... specifying interface + IPv4/v6 NH on some platforms can be useful specifically to stop recursion when the interface goes down by for a nexthop that otherwise may recurse (often to an aggregate or default). Whether it's required to program static routes in i2rs I don't know, but if we plan to support it, a way to stop undesired recursion would be useful. Jon _______________________________________________ i2rs mailing list [email protected] https://www.ietf.org/mailman/listinfo/i2rs
