Moin, I am a co-author of this draft. Enabling communicating an IPv6 nexthop via DHCP enables really interesting setups when combined with an IGP that IPv6 nexthops.
To illustrate the use-case/problem this draft would solve: One can do centralized IPAM for v4 that follows v6 GUA around. Multiple IPv6 networks can receive v4 addressing from a single, central, pool. Think, for example, providing v4 as an add-on for virtual machines/dedicated servers. The ability to communicate a v6 nexthop/default route is the only piece missing to realize this in practice. I sketched this setup out in a RIPE88 talk about V4LESS-AS: https://ripe88.ripe.net/archives/video/1358 https://ripe88.ripe.net/wp-content/uploads/presentations/15-ripe_88_v6_wg.pdf Using the ISC DHCPv4 over DHCPv6 setup and a rather static patch by equi that presumes a v4 default route i setup such a scenario already. In that setup, the lease file gets then re-injected into the IGP based on inotify on the lease files (this-should-be-done-better implementation ;-)). This allows me to have: - One network segment handing out PD from 2001:db8:A::/48 in location A (via a local relay speaking only v6 unicast to the DHCP server) - One network segment handing out PD from 2001:db8:B::/48 in location B (via a local relay speaking only v6 unicast to the DHCP server) - A DHCP server on 2001:db8:D::dc6 in a third location handing out addresses from clients in 192.0.2.0/24 to v6 clients in 2001:db8:A::/48 and 2001:db8:B::/48 Despite the rather flaky setup, this actually works (-actually setting a nexthop based on the DHCP message, but instead reusing the v6 default gw.) With best regards, Tobias On Mon, 2024-10-21 at 23:08 +0200, David 'equinox' Lamparter wrote: > Hi all, > > > there's a fresh new -00 suggesting a new DHCPv4 option to carry an > IPv6 > nexthop for an IPv4 route (for use with RFC5549 style setups, also > cf. > draft-chroboczek-intarea-v4-via-v6-01). > > This is intended to reduce IPv4 wastage when assigning individual > IPv4 > addresses as /32s, by removing the need for the router(s) to have an > IPv4 address. If DHCP4o6 is used, the DHCP server/relay doesn't need > an > IPv4 address either, allowing the end host to be the only system on a > broadcast domain that has an IPv4 address. > > Cheers, > > > equi > > > ----- Forwarded message from internet-dra...@ietf.org ----- > Subject: New Version Notification for draft-equinox-intarea-dhcpv4- > route4via6-00.txt > Date: Mon, 21 Oct 2024 09:53:56 -0700 > > Name: draft-equinox-intarea-dhcpv4-route4via6 > Revision: 00 > Title: DHCPv4 Option for IPv4 routes with IPv6 nexthops > Date: 2024-10-21 > Group: Individual Submission > Pages: 8 > URL: > https://www.ietf.org/archive/id/draft-equinox-intarea-dhcpv4-route4via6-00.txt > Status: > https://datatracker.ietf.org/doc/draft-equinox-intarea-dhcpv4-route4via6/ > HTML: > https://www.ietf.org/archive/id/draft-equinox-intarea-dhcpv4-route4via6-00.html > HTMLized: > https://datatracker.ietf.org/doc/html/draft-equinox-intarea-dhcpv4-route4via6 > > > Abstract: > > // This draft lives at https://github.com/eqvinox/dhc-route4via6 > > As a result of the shortage of IPv4 addresses, installations are > increasingly recovering IPv4 addresses from uses where they are > not > strictly necessary. One such situation is in establishing next > hops > for IPv4 routes, replacing this use with IPv6 addresses. This > document describes how to provision DHCP-configured hosts with > their > routes in such a situation. -- Dr.-Ing. Tobias Fiebig T +31 616 80 98 99 M tob...@fiebig.nl _______________________________________________ Int-area mailing list -- int-area@ietf.org To unsubscribe send an email to int-area-le...@ietf.org
