Re: [v6ops] draft-moreiras-v6ops-rfc3849bis-00

[Owen DeLong <[e-mail]>]

>> I disagree. See my previous message. It has helped in IPv4 and I see
>> no reason it would not help in IPv6.
> 
> Sorry Owen, you need to apply IPv6 think, in particular ULA think
> to the problem space which is different to both IPv4 think and
> general IPv6 think.

Stating that without specifics is useless.

>> It may not do much to prevent mistakes, but it makes them much easier
>> to identify and resolve.
> 
> Copying a documentation prefix is NOT a error.

If you use it in a real world network, yes, it is.

> And I'm sure you will see routes leaked for fc00::/7 as well.  Those
> leaks do not in general cause problem for anyone but the leakers
> themselves.  The fact that people leak has nothing to do with
> documentation prefixes.

It is proof that people do harmful things where the harm is easier to
identify and mitigate if they are constrained to obviously visible and
different from normal numbers.

>>> All a documentation prefix will do is have one potentially toxic
>>> prefix compared to a handful of potentially toxic prefixes.  There
>>> are 1099511627776 prefixes in fd00::/8.  Even if there were thousands
>>> of documentation prefixes the odds of you hitting one when generating
>>> your own prefix are astronomically small.  Additionally the more
>>> prefixes that are used in documentation the less toxic a particular
>>> prefix will be.
>> 
>> No, it will also help to constrain such errors to an easily identifiable
>> single
>> prefix which will make the errors easier to identify and correct. This
>> is, IMHO,
>> a worth while goal in the real world where operators actually run networks
>> and have to deal with less than ideally trained predecessors and
>> colleagues.
> 
> So a site has already choosen a ULA prefix that happens to match
> the soon to be documention prefix.  They have done NOTHING wrong.
> It is causing NO operational problem.  By creating a documentation
> prefix and telling them it is a poision prefix you have just forced
> them to renumber the moment they attempt to connect to anyone.

If we choose the doc prefix from fc00::/8 and not from fd00::/8, then, that
is not true. Since there is currently no accepted valid use for fc00::/8,
and that space is intended to be globally coordinated if it is ever released,
using it to create the doc prefix seems an entirely logical choice to me.

> You are trying to solve a non existent problem.  All you are doing
> is saving the occasional site that was too stupid to pick a prefix
> for themselves and copied one from some documentation from renumbering
> when they connected to another site that was equally stupid and
> copied from the same documention examples when ULA doesn't say you
> won't have to renumber when you connect and exchange ULA routes.

The problem exists, whether you accept that or not. Constraining the
problem is useful to some and harmful to none.

Unless you can show harm caused by allocating a prefix from fc00::/8
for documentation purposes, then I don't see the point of your continued
argument.

Owen