Extending the Vector Packet Processing Engine
I've been building core networking components to leverage VPP more fully as a branch router. Here is an overview of that work.https://enigmatick.social/objects?uuid=b5cfe32e-e1ba-40da-80a1-e6f5bcfb6149

This document discusses a deployment scenario called "an IPv6-mostly network", when IPv6-only and IPv4-enabled endpoints coexist on the same network (network segment, VLAN, SSID etc). The proposed approach enables smooth and incremental transition from dual-stack to IPv6-only network by allowing IPv6-capable devices to remain IPv6-only while the network is seamlessly supplying IPv4 to those that require it.
@MarkTwoFive
If your device can run without IPv4 it will not assign an IPv4 address.
The network provides a NAT64 Gateway to reach the legacy internet
e.g. MacOS, iOS, Android, Dev Version of Linux with NetworkManager.
Extending the Vector Packet Processing Engine
I've been building core networking components to leverage VPP more fully as a branch router. Here is an overview of that work.https://enigmatick.social/objects?uuid=b5cfe32e-e1ba-40da-80a1-e6f5bcfb6149
The Upstream for Ubuntu is not updated yet.
I would spend time to get this update
Unfortunately #Debian contributions work so much differently to what I'm used to today.
@Zugschlus could you help to get started with Debian Maintenance?
I like the idea of global anycast NAT64 gateways.
Maybe we shouldn't invent another prefix for it.
https://datatracker.ietf.org/doc/draft-matolin-global-nat64-anycast/
This document defines a globally routable, anycast NAT64 service using the IPv6 prefix 2600:6464::/96 as a standardized translation substrate for IPv6-to-IPv4 connectivity. The goal of this specification is to eliminate per-network NAT64 configuration complexity by introducing a single globally consistent NAT64 translation prefix operated as a distributed anycast service by participating Internet Service Providers, cloud providers, and content delivery networks. The model assumes an IPv6-only client environment with mandatory IPv4 reachability via NAT64 translation. IPv4-only services remain reachable without modification. IPv4 is not modified. IPv6 is not modified. Only translation placement and routing semantics are standardized. This document defines: * A globally shared NAT64 prefix (2600:6464::/96) * Anycast-based NAT64 edge behavior * Stateless IPv6-to-IPv4 synthesis rules * Optional reverse mapping constraints (IPv4->IPv6 blocked) * Operational requirements for participating networks
Just got a ping about ipxlat being proposed to the linux kernel (in kernel and mainline NAT64/NAT46):
https://lore.kernel.org/netdev/2026031[email protected]/T/#m6b6603d690f9bd7f85cce6307be7821100d98de0