[AusNOG] Prediction: Google et. al. may artificially penalise IPv4 clients

[James Andrewartha]

On Tue, 2 May 2017, Mark Smith wrote:

> Why do you think DHCPv6 support is important to enterprise networks?
> What do they need it for?

I will defer to the admins who posted in 
https://issuetracker.google.com/issues/36949085 but in short, they made 
the choice, the RFCs support them (if you ignore the RFC co-written by the 
Android developer to support his own position), they're not going to 
change their minds. If nothing else, refusing to support DHCPv6 prevents 
implementing DHCPv6-PD which would help with the very feature (USB 
tethering) he says will break if he implements DHCPv6. 464xlat is the 
other thing that won't work, but IPv4 translation is for the network 
provider to manage anyway.

OK, I read the thread for you, and one notes that Cisco only recently 
added RDNSS support, but it won't make it to older hardware. Device 
tracking is another, keeping ND caches is not a good solution compared to 
DHCP logs. Another post "the network admin wishes to apply DNS updates, 
modify firewall rules, etc. when a lease is established and therefore 
chooses stateful addressing via DHCPv6; the network offers additional 
information/services beyond trivial RAs (e.g., time zone, NTP servers, 
printers, PXE boot, SMTP servers, etc.), and therefore sets the O bit in 
RAs"

Please note: I do not run one of these networks, my interest is more in 
pointing out that the decisions of a single man can greatly affect IPv6 
rollout across the world and there's basically nothing you or I can do 
about it.

-- 
# TRS-80              trs80(a)ucc.gu.uwa.edu.au #/ "Otherwise Bub here will do \
# UCC Wheel Member     http://trs80.ucc.asn.au/ #|  what squirrels do best     |
[ "There's nobody getting rich writing          ]|  -- Collect and hide your   |
[  software that I know of" -- Bill Gates, 1980 ]\  nuts." -- Acid Reflux #231 /