Re: getaddrinfo address ordering [Re: IPv6 transition architecture discussion]

[Keith Moore <moore@cs.utk.edu>]

> > at least not right now.  attempts to talk to v6 sites typically
> > time out (which is annoying because the v6 addresses are tried first -
> > another thing which needs to be fixed)
> 
> Do I sense this as a voice for support to modify or at least explore the
> current de-facto standard ordering?

I'm certainly going to change it on the machines that I use
(it helps that I have source code)
 
> > it turns out that you need the address ordering to be sensitive to both
> > network configuration and to the particular application.  some apps will
> > be v4 by default, others will be v6 default, others will be exclusively
> > one or the other.
> 
> There are already some variables that affect this, namely:
> 
>  1) whether the DNS lookup produces A, AAAA, or both; ie "remote
> site configuration"

I don't see how this affects ordering - if you don't get back A 
records are you going to put the AAAA records in a different
order relative to one another than if you do get them back?
 
>  2) what ordering for DNS records is used when both are used
> (default is AAAA first everywhere I know, no possibility to change that)
> 
>  3) whether the application uses AF_INET, AF_INET6 or AF_UNSPEC
> (with server apps, this is even a bit muddier.)
> 
> These help quite a bit, but I guess adding some getaddrinfo hint like
> AI_PREFERV4 or AI_PREFERV6 could be added in the case that DNS lookup
> returns both addresses, you use AF_UNSPEC and you want to influence the
> default DNS record ordering.

it's fairly easy for an app to reorder addresses that it gets back
between IPv4 and IPv6.  but I often find myself wanting to specify
a mixed preference like:

6to4 addresses
native IPv4
other IPv6

Keith