Re: [v6ops] I-D Action: draft-colitti-v6ops-host-addr-availability-01.txt

[Mark Smith <[e-mail]>]

On 29 July 2015 at 02:51, Lorenzo Colitti <[e-mail]> wrote:
> On Wed, Jul 29, 2015 at 1:28 AM, Ted Lemon <[e-mail]> wrote:
>>
>> Do you buy my theory about just using a /120 to aggregate a bunch of
>> addresses, or no?
>
>
> Well but if if you give a laptop or a device a /120, what is it going to
> give its downstream devices / VMs when sharing its internet connection with
> them? Does it have to implement a stateful DHCPv6 server and force its
> downstream devices/VMs to use it? At that point, the network it offered
> would be violating recommendations made in the draft.
>

For a laptop style scenario, i.e., numbers of VMs in likely low 1s or
no more than 10 to 15, I'd think bridging them onto the existing
ethernet network is easier because it is using existing and operating
SLAAC, DHCP etc. That has been my experience on the few occasions I've
run up a VM or two. It also has the advantage that it doesn't require
anything special from the upstream network that you might not be an
operator of.

I think that means there is a threshold, below which bridging is most
efficient, above which routing via a DHCPv6-PD delegated prefixes is
most efficient.

<snip>

Regards,
Mark.