Re: [Gen-art] Genart telechat review of draft-ietf-bfd-vxlan-09

[Greg Mirsky <[e-mail]>]

Hi Erik,
thank you for your reviews and for sharing thoughts on the selection of the
destination IPv6 address. Following recommendations from Adam, I've updated
the document to use the proper representation of IPv6 addresses and refer
to them as "IPv4-mapped IPv4 loopback addresses". These updates are in the
attached diff. Adam also noted that RFC 8504 doesn't have a similar wording
regarding the handling of packets addressed to an address from 127/8
network as RFC 1812 (of course, referring to IPv4-mapped 127/8 addresses):
      A router SHOULD NOT forward, except over a loopback interface, any
      packet that has a destination address on network 127.  A router
      MAY have a switch that allows the network manager to disable these
      checks.  If such a switch is provided, it MUST default to
      performing the checks.
I'd note, that the egress BFD system is expected to accept a BFD packet
with the destination IP address from the specified range without being
provisioned for the specific address from that range. Perhaps that makes
the use of this range possible even though its special handling is not
explicitly documented.

Best regards,
Greg


On Mon, Dec 16, 2019 at 9:19 PM Erik Kline via Datatracker <[e-mail]>
wrote:

> Reviewer: Erik Kline
> Review result: Ready with Nits
>
> I am the assigned Gen-ART reviewer for this draft. The General Area
> Review Team (Gen-ART) reviews all IETF documents being processed
> by the IESG for the IETF Chair. Please wait for direction from your
> document shepherd or AD before posting a new version of the draft.
>
> For more information, please see the FAQ at
>
> <https://trac.ietf.org/trac/gen/wiki/GenArtfaq>.
>
> Document: draft-ietf-bfd-vxlan-??
> Reviewer: Erik Kline
> Review Date: 2019-12-16
> IETF LC End Date: None
> IESG Telechat date: 2019-12-19
>
> Summary:
>
> -09 addresses my concerns from -07.  Thank you for this.
>
> The one "nit" is that it seems to have introduced a recommendation to use
> ::ffff:7f00:0/104 as an IPv6 loopback prefix.  (a) This document should
> follow
> the format recommendations of RFC 5952 section 4.3 and lowercase the
> "F"s.  But
> (b) more importantly, I'm not sure how implementations may treats this
> space.
>
> The use of an RFC4291 section-2.5.5.2 mapped v4 address doesn't necessarily
> make the packet a part of an IPv6 connection.  Nevertheless, I'm not sure I
> have a strong feeling about this as it may still exercise enough of the
> IPv6
> stack in a VTEP.
>
> I definitely do think that in the case of BFD on the management VNI
> targeting
> an IPv6 link-local address of the VTEP would be better.  However, I expect
> that
> if ::ffff:127.0.0.0 does prove to have some issues in the future a -bis
> can be
> written quickly with a recommendation.
>
> Also, Suresh may have ideas for a solution.
>
> Major issues:
>
> Minor issues:
>
> Nits/editorial comments:
>
>
>