Android (lack of) support for DHCPv6

Karl Auer kauer at biplane.com.au
Wed Jun 10 11:30:40 UTC 2015


On Wed, 2015-06-10 at 19:49 +0900, Lorenzo Colitti wrote:
> Question for everyone on this thread that has said that DHCPv6 NA is a
> requirement: suppose that Android supported stateful DHCPv6 addressing,
> requested a number of addresses, and did not use any of them if the number
> of addresses received was less than N.
> What does N need to be?

I think that's a wrong question, or maybe I am completely missing your
point.

Seems to me that N will vary depending on what you are trying to do. And
you could well be trying to do several things at once, each with a
different requirement. And these things may happen over time, so that at
one time you need N, while later you need ten times that many - or half
as many, or none. With DHCP you just ask for more when you need more, or
release ones you don't need. You don't have to arrange everything up
front and then be stuck with it.

You know how many addresses you need to provide a given service; you
know how to degrade the service gracefully, or whether a graceful
degrade is even possible. In other words, you the requester know how
many addresses you want and how many you have to have - which are two
possibly quite different numbers.

Addresses are just a resource, and like any other resource, if the
environment can't supply them, you either degrade the service, fail to
provide it, or possibly keep trying and provide it later when the
resource becomes available. At their most basic, standard DHCPv4 and
DHCPv6 clients do exactly that - they keep trying until they get
addresses.

Not being able to get enough addresses is pretty much like not being
able to get enough RAM or disk space, but you make it sound like running
out of power!

It is not an all-or-nothing proposition at a platform level, and
demanding to know "what is N?" implies that it is. 

Regards, K.

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Karl Auer (kauer at biplane.com.au)
http://www.biplane.com.au/kauer
http://twitter.com/kauer389

GPG fingerprint: 3C41 82BE A9E7 99A1 B931 5AE7 7638 0147 2C3C 2AC4
Old fingerprint: EC67 61E2 C2F6 EB55 884B E129 072B 0AF0 72AA 9882





More information about the NANOG mailing list