Greetings all,
</chair hat>
I've re-reviewed this version of the draft, and think that, on balance, it is a
useful draft explaining how an operator may utilize IP VRF based VPN's to ease
the deployment o NAT44 based CGN's. There are a few suggestions for the
authors, outlined below.
1) I would indicate that the use of a services VRF is a MAY (in IETF parlance).
I could have the services address space either in a services VRF, or I could
actually have it in my non VRF tables (reducing the number of VRFs that I have
to maintain in the network).
2) There is an assumption that appears in the document that the use of CGN's
will increase over-time, leading to a decentralized model. I would instead
phrase it that the model could be centralized or decentralized based on
architectural, operational, or engineering requirements. As far as CGN use
growth, I would phrase it that CGN deployment will probably follow a bell-like
curve. It grows as I am adding more IPv4 customers when I have exhausted my
IPv4 public block, and then will shrink as IPv6-only or the other transition
mechanisms take hold. Forecasting a continual growth of CGN is probably
(hopefully) incorrect.
Christopher
_______________________________________________
OPSAWG mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/opsawg
