Hello John, and thanks for the comments. > But the rub is that every node of that IPv6-only core network (in this > vision) nevertheless must know about IPv4 routing. [...] In other > words, the draft does not envision deployment in an IPv6-only network of > IPv6-only routers.
You're entirely right. We're assuming a situation in which essentially all routers implement IPv6, but some networks still haven't deployed IPv6 addresses. I think it fairly uncontroversial that that's the situation we're in today. > How many networks are there like this, really? Networks where routers support both IPv4 and IPv6, but where IPv6 hasn't been deployed yet? Quite a few. > So let's now look at how the vision would actually work today. > > Let's assume that there is a route for 123.0.0.1/24 whose next hop is > 2023:0124::0666, which is on interface eth7. Does a packet for > 123.1.2.3 get sent out as an IPv4-over-Ethernet packet on eth7? How > could it be addressed to 2023:0124::0666 as an IPv4-over-Ethernet > packet? It's addressed to 123.0.0.1. The IPv6 address of the next hop is only used for MAC address resolution, and never appears on the wire. This is not different from what happens in ordinary next-hop routing. > The document is silent on all these details. I most respectfully disagree. Section 1 describes the procedure in detail. Section 3.2 makes it clear that the only change required in router implementations is the ability to carry both v4 and v6 next hops in the IPv6 routing table. > It seems like the idea is to save money by not requiring increasingly > expensive IPv4 addresses on every router interface, Not just money (most folks are happy with 10/8, which is free), but administrative overhead. It avoids adding IPv4 addresses to all routers and ensuring that they are unique. > but instead to push the costs onto the rest of the Internet, which would > have to upgrade their software to support some kind of odd change to ICMP. There are no changes to ICMP. The generated packets are perfectly ordinary ICMPv4 packets, they just happen to have a kinky source address. > It also looks like something like this would probably work as long as > there is a single unique IPv4 address assigned to each router (not one > IPv4 address per interface). That address could be used for ICMP > replies on any interface. Does that work? Yes, it does, and it's been implemented in most routers for decades (with slightly different behaviour depending on the router vendor). It's a good alternative to v4-via-v6 routing in circumstances where assigining IPv4 addresses to routers is not prohibitive. > PS: The draft seems to confuse IPv6 link-local addresses (FE8x::, RFC > 3513) with IPv6 unique local addresses (FDxx::, RFC 4193). I'm reasonably sure it doesn't. Please explain what makes you think that. Thanks, -- Juliusz _______________________________________________ Int-area mailing list -- [email protected] To unsubscribe send an email to [email protected]
