[lisp] Changes that have demonstrated sufficient consensus

[Sam Hartman <[e-mail]>]

>>>>> "Dino" == Dino Farinacci <[e-mail]> writes:

For the changes listed below, it's fine to submit, or to wait until we
get closer on some of the open issues.

Note that the issue tracking policy we agreed to requires there to be
a change log section in the draft, not just in mail to the list.  I
don't see that in 04; that will need to be added before it is
submitted.

The level of detail for the changes below is fine.


    Dino> * Add Fred Templin in ack section.

I certainly don't object to you calling this out in the changelog, but you need not do so.
    Dino> * Say more about LAGs in the UDP section per Sam Hartman's
    Dino> comment.

According to the issue tracking policy, you should have followed up with me.
However I've reviewed the text and it is fine; thanks.

    Dino> * Sam wants to use MAY instead of SHOULD for ignoring
    Dino> checkums on ETR.  From the mailing list:

    Dino>   You'd need to word it as an ITR MAY send a zero checksum,
    Dino> an ETR MUST accept a 0 checksum and MAY ignore the checksum
    Dino> completely.  And of course we'd need to confirm that can
    Dino> actually be implemented.  In particular, hardware that
    Dino> verifies UDP checksums on receive needs to be checked to
    Dino> make sure it permits 0 checksums.

    Dino> * Margaret wants a reference to
    Dino> http://www.ietf.org/id/draft-eubanks- chimento-6man-00.txt.

I think we're still waiting for Joel and Margaret to give us final wording on the UDP checksum text.
Feel free to include what is in 04: it is clearly closer to WG consensus than 03.

    Dino> * Fix description in Map-Request section. Where we describe
    Dino> Map-Reply Record, change "R-bit" to "M-bit".

This may be small enough that it doesn't need to be called out.

    Dino> * Indicate SHA1 can be used as well for Map-Registers.

Note that the text is still broken.  The IPsec sha-1  is 96-bits long not 128-bits long.
Also, according to some software I'm using but not verified with the spec, 0 is an illegal SPI.
However this change is small and is supported by WG discussion.

    Dino> * More Fred comments on MTU handling.

    Dino> * Isidor comment about specing better periodic
    Dino> Map-Registers. Will be fixed in draft-ietf-lisp-ms-02.txt.

    Dino> * Change "loc-reach-bits" to "loc-status-bits" per comment
    Dino> from Noel.

    Dino> * Clarify that when E-bit is 0, the nonce field can be an
    Dino> echoed nonce or a random nonce. Comment from Jesper.

    Dino> * Indicate when doing data-gleaning that a verifying
    Dino> Map-Request is sent to the source-EID of the gleaned data
    Dino> packet so we can avoid map- cache corruption by a 3rd
    Dino> party. Comment from Pedro.

Fine if the gleaning text in general is OK.

    Dino> * Indicate that a verifying Map-Request, for accepting
    Dino> mapping data, should be sent over the the ALT (or to the
    Dino> EID).

    Dino> * Reference IPsec RFC 4302. Comment from Sam and Brian Weis.

Note this probably has bigger implications than you expect; it moves
you from 2401 to 4301 for the base architecture and that has huge
implications.

    Dino> * Put E-bit in Map-Reply to tell ITRs that the ETR supports
    Dino> echo- noncing.  Comment by Pedro and Dino.

This might have needed an explicit wg list message, but I'm going to
let it slide: it's a really obviously useful idea.

    Dino> * Jesper made a comment to loosen the language about
    Dino> requiring the copy of inner TTL to outer TTL since the text
    Dino> to get mixed-AF traceroute to work would violate the "MUST"
    Dino> clause. Changed from MUST to SHOULD in section 5.3.