Description
Details
- Version
- -
- Is it a breaking change?
- Unspecified (possibly destroys the router)
Event Timeline
Here is an old implementation for Vyatta. Moreover Jool does MAP-T, offers a kernel module, and is being implemented into nftables.
While MAP is not an urgent feature I think it would a good idea for VyOS to be ahead of the curve on this one, as transition methods are gaining quite a lot of traction (see the RIPE Open House).
I just want to add that my ISP is dropping dual stack in next few months and at the end of 2021 all users will only have IPV6 networking capabilities.
For thoose who still require IPV4 traffic it will be "tunneled" via MAP-T.
Right now there are close to 0 CPEs able to do this. The only open source solution is OpenWRT.
Even professional solutions are lacking this feature.
Just wanted to add support for MAP-T/E since this is a core IPv6 transition mechanism for new ISPs that host v6 only networks. Competing projects like FDio/VPP have this in their core, so it would be great to see in the XDP / VyOS platform.
The cernet MAP latest commit was 10 years ago.
We have VPP as an addon https://blog.vyos.io/vyos-and-vpp-progress-and-plans
Hey @syncer is this feature being bumped to a future release? I saw the Project tags get swapped to a GA tag (which sounds more official to me) but also a comment of "no business case."
Either way, I'm hopeful this feature will get prioritized. This is a major need to keep VyOS competitive in an IPv6-first world. We are an ISP that runs native v6 and MAP on our Border Routers - VyOS is not in the running without this feature.
As mentioned previously there are lots of implementation examples with projects like VPP, and since the VyOS is starting to implement that in the data plane, this feature should be a relatively easy control plane addition when that happens.
Thoughts?