IPv6 Ignorance

[Benedikt Stockebrand]

Hi everybody,

Tim Densmore <tdensmore at tarpit.cybermesa.com> writes:

> While I posted in that thread in favor of IPv6 Deployment, I have to
> admit that I'm somewhat sympathetic to their views.  Aside from being
> future-proof, what real purpose does being IPv6 "ready" do for a
> network/network operator?
> [...]

the economic impact of an IPv6 deployment can be extensive indeed.
Aside from the more or less obvious pressure on management---to make
the most of whatever company-wide resources they have---the question
about the "use" of IPv6 is often difficult to answer.

IPv6 doesn't have a "business case" in the normal sense, but there
*is* a line of reasoning that meets management logic: "As soon as IPv6
has been made available to the people who by their sheer numbers
finance the Internet, i.e. the private end users, either you also
support IPv6 or you are effectively disconnected from the Internet.
When that has happened you will need to survive with a significantly
reduced functionality especially when communicating with the outside,
like your customers and suppliers, until your---likely overworked and
underskilled IT department---has caught up."


If I get through with this line of reasoning---which sometimes
actually happens---then I suggest a number of low key but long term
measures to deal with IPv6 in a company environment:

- Only buy IPv6-capable products (use RIPE-501bis if you don't
  know what exactly to watch out for).  Not doing so will lead to
  significant unbudgeted emergency investments later on instead of
  requiring little extra money and only a modicum of extra thought
  right now.  @Tim: This could have saved you from getting stuck with
  switches not supporting RA guard, if somebody had realized about
  five years ago.

- Try to get the techies up to speed with IPv6.  This involves a bit
  of money and time, but you'll have to put that in sooner or later
  anyway.  This affects liquidity, but usually not in an overly
  painful way.  On the other hand it helps to avoid decisions that
  turn against you once you actually need IPv6.

- Either don't make long term contracts with your ISPs, IT
  contractors and such unless they can provide IPv6, or add an "exit
  clause" (proper English term?) in case they can't deliver by a fixed
  date.  Don't forget to include equivalent SLAs.

- Whenever you touch anything in your IT, see if you can get it IPv6
  enabled along the way.  This involves code updates (moving from
  gethostbyname() to getaddrinfo() etc.), hardware replacements, new
  VLANs on your existing network and just about anything.  This is a
  low key approach provided people are familiar with IPv6.

All of these, and a number more, are generally feasible if management
is at all willing to accept that they need a working IT not only today
but two years from now.  The alternative is being overrun by the
eventual rush towards IPv6, resulting in expensive crisis projects
that involve lots of unnecessary risks, usually end with a rather
suboptimal result and give your competitors a time window to increase
their market share.

Or, doing consulting myself, I simply point out that while I might be
available to them by the time IPv6 hits the Internet large scale, due
to the increased demand my rates won't be the ones I charge today.
Again, that's a sort of reasoning any MBA should understand.


Cheers,

    Benedikt

-- 
			 Business Grade IPv6
		    Consulting, Training, Projects

Benedikt Stockebrand, Dipl.-Inform.   http://www.benedikt-stockebrand.de/