Re: IPv10 (Temp. name IPmix) (draft-omar-ipv10-00.txt).

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 01/01/2017 13:25, Randy Bush wrote:
>>>> Android users can complain to whoever it is that supports Android.
>>> for some silly reason, my customers don't think they pay me for
>>> blame shifting.  they just want things to work.
>> Yes. So not implementing DHCPv6 might be a self-defeating decision
>> by an operating system developer, don't you think?
> 
> the developer in question is an ipv6 purist.  shooting himself in the
> foot and the customers at the same time.  it's an ipv6 tradition.
> 
>>> and, let me repeat for the fourth time, enterprises of scale use dhcp
>>> to drive clients to the desired exit.
>> Excuse my ignorance, but which DHCP options does that involve?
> 
> code 3 in v4, router option.  folk use it to cause subsets of the space
> to take different exits.

Thanks, I was wondering whether you meant something more subtle than that.
And yes, things like draft-ietf-mif-dhcpv6-route-option and
draft-sarikaya-dhc-dhcpv6-raoptions-sadr have a sad history of being trampled.

In fact I haven't changed my mind since 2012:
https://mailarchive.ietf.org/arch/msg/mif/B6NCyw2y2O6I112TQJp_pcN_OoM

>>> dhcp6 does not let them do that.  without feature parity we don't get
>>> to play.
>> Which is why I think RFC 8028 has value.  It's intended to support
>> exit selection via first-hop selection.
> 
> that is not feature parity.  that is yet one more ipv6 'feature' that is
> just different for religious reasons.  

It's a solution that will work the same for SLAAC-derived and
DHCPv6-derived host addresses (and for manually assigned host addresses,
for that matter), so I don't think that is very fair. I agree that it
moves the necessary config magic from the DHCP(v6) screen to the router
screen, but the logical result is the same. Given the lemons available,
it seems like a reasonable lemonade recipe.

However (see above) I baiscally agree with you. When MIF was first proposed
I said "all we really need is a default router per prefix" and IMNSHO that's
still true.

> instead of giving the customer
> what they want, we invent shiny new stuff and wonder why they walked
> away.  it must be that they are stupid.

Hardly, but they are sometimes a bit set in their ways. Change costs
money.

    Brian




[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]