Re: [dnsext] Fwd: New Version Notification for draft-ah-dnsext-rfc1995bis-ixfr-02

[Edward Lewis <[e-mail]>]

At 16:15 +0200 6/17/11, OndÞej Sur˜ wrote:

>But the fact is that you get a full zone.
>
>Also the new draft says:
>
>    In case of fallback to AXFR, the answer contains the same RRs (and is
>    subject to the same ordering rules) as specified in Sections 2.2
>    through 3 of RFC 5936, with the single difference being the echoed
>    QCODE (i.e., IXFR) in the Question section.
>
>The fact is that the it's the behaviour which is "AXFR" (that is what we
>talk about), but the QCODE/RCODE is not "AXFR", but IXFR.

My impression is that the issue surrounding 
IXFR-only is about load on processing and network 
latency.  I fail to see that the problem is that 
you get all of the zone at once. The problem is 
that you have to do a lot of work to get all of 
the records of an large zone when all you really 
need is a just a small increment.

Why else would IXFR-only be needed?

Is it just to discover all of the increments? 
That sounds like a poor reason.  The increments 
aren't important except to satisfy onward IXFR 
answers and even then, if the zone is small, why 
not just use AXFR-style IXFR?

>It is optimization (look into archive, we talked about that before and
>it should be written in Motivations for IXFR-only section of the I-D).

The archives are not searchable - have links?
-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis
NeuStar                    You can leave a voice message at [phone]

I'm overly entertained.