[AusNOG] IPv6 - What Should an Engineer Address when 'Selling' IPv6 to Executives?

[Mark Andrews]

In message <513689AB.8080508 at libertysys.com.au>, Paul Gear writes:
> Respectfully, Robert - you have not addressed Don's question.  You've 
> made an excellent case for what /not/ to do (and i wholeheartedly 
> agree), but nothing on the side of the case for what to /do/.
> 
> I'd really be interested in hearing the group's take on the actual 
> benefits of IPv6 /now/.  The only one i can think of is:
> 
>   * reduced network complexity and support costs due to the elimination
>     of NAT

You won't remove NAT for IPv4 for many years to come.  You are going
dual stack not IPv6-only.

The biggest advantage of moving now is that you spread your costs
over time and can backout if you experience problems you can't over
come otherwise.  You also get to workout what needs to be upgraded
/ replaced before it gets on the critical path.  The lead times for
that can be months if not years.  You get to test the IPv6 support
in your systems when you are not under presure.

Dipping your toe into IPv6 is not expensive.  It's usually no more
that configuring a border firewall to be reply traffic only, turning
on the upstream and enabling RAs.  You can stop and test after each
of these steps.  If something goes wrong you back out that step.

Mark

> I expect that this benefit would be eliminated many times over by the 
> cost of:
> 
>   * implementing IPv6 operationally on the network
>   * retraining staff (I can't even get the help desk folks to stop using
>     IPv4 addresses when DNS is already set up and working!)
>   * re-implementing firewalls to eliminate NAT
>   * (for small multi-homed organisations):
>       o applying for provider-independent address space
>       o implementing BGP (including acquiring the hardware and
>         skills/partners to roll it out)
> 
> All the other benefits of IPv6 that i can think of are future:
> 
>   * ability to grow the network beyond the present limits imposed by
>     IPv4 addressing
>   * not being subject to the increased cost of acquiring more than a /22
>     of IPv4 address space
>   * competitive advantage over competitors who have neglected to plan
>     for the future
>   * not being subject to the exhaustion of IPv4
>   * not being cut off from customers who have already fallen victim to
>     the exhaustion and are IPv6 only
> 
> It would be a foolish manager who ignored that big list of future 
> benefits (especially the last two), but i can very much understand him 
> or her continuing to defer it until next year's budget, or at least 
> pushing the project down the priority queue until bitten.
> 
> IPv6 is a technical /must //do/,//and i continue to be wracked with 
> professional guilt that i didn't start sooner, invest more learning 
> time, and build up more practical experience with it, but i haven't 
> managed to find a silver bullet for convincing management.  Some of the 
> best Internet minds in the Asia-Pacific region are on this list.  If 
> they can't come up with some good sells for the benefits of IPv6 now, 
> what hope does the average IP professional have?
> 
> Regards,
> Paul
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: marka at isc.org