[arin-ppml] The non-deployment of IPv6 - The Economic Factor

Ted Mittelstaedt tedm at ipinc.net
Tue Dec 29 11:39:19 EST 2009

Matthew Keller wrote:
> Sorry to dredge an "old" thread back, but I'm catching up on PPML and 
> Vaughn's note really nailed a large segment of what I've been explaining 
> for several years now when dealing with the "deploy IPv6 or you suck and 
> are a bad netizen" crowd.
>> This (*The Economic Factor*) should not be underestimated
> I'll go one step further - Deploying IPv6 across a perfectly-running 
> enterprise network is excruciatingly daunting.
> This isn't daunting because admins lack the expertise or confidence to 
> do so, or haven't done it a bajillion times in test environments, it's 
> because our customers (and bosses) don't care - they don't care one lick 
> that you're rolling out IPv6, whatever that is. The network is 
> _working_, everything is _fine_ and they very simply don't want to risk 
> something not working anymore: If it ain't broke, don't fix it. The 
> customer has plenty of other things they _do_ want.

They didn't give a rat's ass what they were running on when we
switched them from IPX-only to dual-stack
IP and IPX a decade ago when we were switching from Netware to
Windows Server, and then later switched them from dual-stack
IPX and IP to IP-only.

This is an excuse, pure and simple.  But, it's your network,
run it any way you want.  If you want to do nothing until you
absolutely have to access an IPv6 content provider then slap
in an IPv4->IPv6 gateway at your border, your welcome to
do so.

> There is little-to-no consumer-driver to enable IPv6 in established 
> networks, and when coupled with Vaughn's assessment of economic 
> dissuasion, it paints a picture that unless you are very well-off 
> (capitally and/or human-resourcely), _forced_ by outside powers 
> (something to show your customers and bosses to justify the risk of 
> change), or setting up a new network from scratch (you'd be stupid not 
> to): Trying to get "normal", established edge networks to deploy IPv6 is 
> one-step down from trying to herd cats.

Perhaps.  But I do know one thing - 10 years from now when both you and
I are out of work and looking for a job, my Resume is going to say IPv6
experience and yours won't - and I'll likely get a position faster
and with more compensation than you do.

Suit yourself.


More information about the ARIN-PPML mailing list