Re: [v6ops] An Update to Happy Eyeballs

Hi Nick,

We can clarify the text to make the distinction clearer.

This case refers to the sending of A and AAAA queries to your DNS server:

> Both queries SHOULD be sent on the wire as
>   close together as possible, with the ordering being dictated by the
>   system's host address preference policy.

The second case refers to which DNS server we send to, if we have both an IPv6 address for a DNS server and an IPv4 address for a DNS server:

> In keeping with the Happy Eyeballs approach,
>   queries SHOULD be sent over IPv6 first.

So, to summarize:
- Choose which DNS server to talk to first if you have multiple; prefer IPv6 if you have no better information that's what your system prefers
- When you send queries to that server, send both A and AAAA close together, and if your system prefers IPv6, send the AAAA first.

Thanks,
Tommy 

> On Mar 9, 2017, at 9:22 AM, Nick Chettle <nick@nccnetworks.co.uk> wrote:
> 
> Hi David,
> 
> It looks good, thanks for sharing what you've learnt.
> 
> I just have one question, when talking about the DNS query in section 2 you say:
> 
> Both queries SHOULD be sent on the wire as
>   close together as possible, with the ordering being dictated by the
>   system's host address preference policy.
> 
> Then when talking about DNS server selection in Section 2.1 you mention that:
> 
> In keeping with the Happy Eyeballs approach,
>   queries SHOULD be sent over IPv6 first.
> 
> Are these statements slightly contradictory or should DNS queries relate to the host policy while DNS server selection always uses IPv6 first?
> 
> Thanks, Nick
> 
> 
> 
> 
> 
> ________________________________________
> From: v6ops <v6ops-bounces@ietf.org> on behalf of David Schinazi <dschinazi@apple.com>
> Sent: 08 March 2017 19:25:29
> To: IPv6 Operations
> Subject: [v6ops] An Update to Happy Eyeballs
> 
> Hi v6ops,
> 
> Tommy Pauly and myself have written up what we've learned
> in the past few years maintaining and improving Apple's Happy Eyeballs stack,
> and submitted it as draft-pauly-v6ops-happy-eyeballs-update.
> 
> You'd love to hear your feedback!
> 
> Thanks,
> David Schinazi
> 
> 
> Begin forwarded message:
> 
> From: internet-drafts@ietf.org<mailto:internet-drafts@ietf.org>
> Subject: New Version Notification for draft-pauly-v6ops-happy-eyeballs-update-00.txt
> Date: March 8, 2017 at 10:43:20 PST
> To: Tommy Pauly <tpauly@apple.com<mailto:tpauly@apple.com>>, David Schinazi <dschinazi@apple.com<mailto:dschinazi@apple.com>>
> 
> 
> A new version of I-D, draft-pauly-v6ops-happy-eyeballs-update-00.txt
> has been successfully submitted by David Schinazi and posted to the
> IETF repository.
> 
> Name: draft-pauly-v6ops-happy-eyeballs-update
> Revision: 00
> Title: An Update to Happy Eyeballs
> Document date: 2017-03-08
> Group: Individual Submission
> Pages: 7
> URL:            https://www.ietf.org/internet-drafts/draft-pauly-v6ops-happy-eyeballs-update-00.txt
> Status:         https://datatracker.ietf.org/doc/draft-pauly-v6ops-happy-eyeballs-update/
> Htmlized:       https://tools.ietf.org/html/draft-pauly-v6ops-happy-eyeballs-update-00
> 
> 
> Abstract:
>  "Happy Eyeballs" (RFC6555) is the name for a technique of reducing
>  user-visible delays on dual-stack hosts.  Since one address family
>  (IPv4 or IPv6) may be blocked, broken, or sub-optimal on a network,
>  clients that attempt connections for both address families in
>  parallel have a higher chance of establishing a connection sooner.
>  Now that this approach has been deployed at scale and measured for
>  several years, the algorithm specification can be refined to improve
>  its reliability and generalization.
> 
> 
> 
> 
> Please note that it may take a couple of minutes from the time of submission
> until the htmlized version and diff are available at tools.ietf.org<http://tools.ietf.org>.
> 
> The IETF Secretariat
> 
> 
> _______________________________________________
> v6ops mailing list
> v6ops@ietf.org
> https://www.ietf.org/mailman/listinfo/v6ops