[155624] in North American Network Operators' Group
Re: Return two locations or low TTL [was: DNS caches that support
daemon@ATHENA.MIT.EDU (Nick Hilliard)
Mon Aug 20 10:40:02 2012
X-Envelope-To: nanog@nanog.org
Date: Mon, 20 Aug 2012 15:37:48 +0100
From: Nick Hilliard <nick@foobar.org>
To: "Patrick W. Gilmore" <patrick@ianai.net>
In-Reply-To: <15BF47FF-9B3D-498B-A58B-F03853F0BA08@ianai.net>
Cc: NANOG list <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
On 20/08/2012 14:18, Patrick W. Gilmore wrote:
> On Aug 20, 2012, at 08:47 , Chris Adams <cmadams@hiwaay.net> wrote:
>> Most anything that supports IPv6 should handle this correctly, since
>> getaddrinfo() will return a list of addresses to try.
>
> Ah, the amazing new call which destroys any possibility of randomness or
> round robin or other ways of load balancing between A / AAAA records.
well, "new" as in about 16 years old.
> Yes, all of us returning more than one A / AAAA record are hoping that
> gets widely deployed instantly. Or not.
inet_addr() is zomfg brain damaged, ipv4 only, non-thread-safe and needs to
die horribly in a fire. getaddrinfo() puts at least the possibility of
address selection into the hands of the developer which is going to be
completely necessary for happy eyeballs / rfc3484bis / etc.
Nick
