Facebook over IPv6

jkrejci at usinternet.com jkrejci at usinternet.com
Sat Jun 11 03:18:17 CEST 2011

I'm not condoning or defending their techniques and policies, merely pointing out "wasting rack space in a hosting facility" is not the only solution to meet their requirements. I am sure their reasons for keeping ipv6 quarantined like this is based on the perceived risk of lost ad revenue from that, casually tossed around stat, 0.03% of users who will have a poor or non-functioning v6 experience.

I am guessing they'll switch off this method of v6 quarantining when the percentage of lost user communications (ad revenue) gets to a sufficiently low number or it gets to be too complex and messy to maintain it all. Or maybe they'll just switch on all ipv6 everywhere one day a be a leader among the big names in the internet.

Sent via BlackBerry from T-Mobile

-----Original Message-----
From: Ted Mittelstaedt <tedm at ipinc.net>
Sender: ipv6-ops-bounces+jkrejci=usinternet.com at lists.cluenet.de
Date: Fri, 10 Jun 2011 17:28:55 
To: <jkrejci at usinternet.com>
Cc: <ipv6-ops at lists.cluenet.de>
Subject: Re: Facebook over IPv6

On 6/10/2011 4:28 PM, jkrejci at usinternet.com wrote:
> You do not of course have to run a separate physical server to run a
> separate instance of a name server that does IPv6 only, listening on
> and querying from its own IP address. Unless there are other
> requirements from google. I have not read their DNS whitelisting
> policy in a while.

What technical reason is there that you would run separate IPv4 and
IPv6 DNS servers?  The server software can do it so the only reason
they would stick that requirement in there is redundancy.

And when you virtualize or whatever, you lose the redundancy.  Thus
to me your "cheating" and going against the meaning of the Google

I may tolerate this "the real speed limit is 65Mph but we will
post it 55Mph" on the nation's highways bullshit, but I'm not
going to support it with this.

Going with Google's requirements is just perpetuating some ass-backward
network admins idea of how to transition.  I'd rather bitch about it
and maybe Google will pull their head out if enough other people bitch
too.  The alternative is for every admin to do whatever weird cockamamie
different requirements that a bunch of service providers dream up to
get IPv6 running.

So when Facebook comes out with a requirement that you have a dual-stack
DNS server to get on "their" IPv6 whitelist, are you going to meekly 
accept that too?  How exactly are you going to make both of them happy?


> Sent via BlackBerry from T-Mobile
> -----Original Message----- From: Ted Mittelstaedt<tedm at ipinc.net>
> Sender: ipv6-ops-bounces+jkrejci=usinternet.com at lists.cluenet.de
> Date: Fri, 10 Jun 2011 15:47:25 To:<ipv6-ops at lists.cluenet.de>
> Subject: Re: Facebook over IPv6
> On 6/10/2011 6:19 AM, Bill Owens wrote:
>> On Thu, Jun 09, 2011 at 09:43:00PM -0700, Ted Mittelstaedt wrote:
>>> We never raised a finger to get on the "Google whitelist" but we
>>> have no trouble accessing Google via IPv6.  And we have our own
>>> ARIN-assigned IPv6 block.
>>> IMHO the whole concept of whitelisting IPv6 blocks is
>>> ridiculous.
>> Perhaps, but that's not how the Google DNS whitelist works:
>> http://www.google.com/intl/en/ipv6/
>> The important question is this: what caching resolver do you use
>> to query for google.com names? That's what determines whether
>> you're in the whitelist or not.
>> Bill.
> OK I looked at that, the issue is I was hitting ipv6.google.com not
> www.google.com.  I guess they are only blacklisting you if you
> attempt to pull AAAA records for www.google.com
> Very likely I will never request to be on the whitelist, because of
> the Google stipulation:
> •Separate DNS servers for your IPv6 users (not shared with IPv4-only
> users)
> which is, IMHO, asinine.  We pay a damn lot of money for our telco
> space and I'm not going to waste 2U plus electricity on 2 machines to
> loaf around answering the occasional IPv6 DNS query when we already
> have perfectly good DNS servers.
> What the hell is wrong with these assholes?  Who the hell do they
> think they are?
> Ted

More information about the ipv6-ops mailing list