Re: [v6ops] RFC6459 "IPv6 in 3GPP" - the IID in the LL address

[Alexandre Petrescu <[e-mail]>]

Gert,

Le 10/07/2017 à 19:59, Gert Doering a écrit :
> Hi,
> 
> On Mon, Jul 10, 2017 at 07:53:52PM +0200, Alexandre Petrescu wrote:
>>> [..]
>>>> Some packets with LLs as src have been witnessed in the Internet at large.
>>>
>>> This is a bug in all the forwarding entities on the path and needs to be
>>> fixed.
>>
>> That is your oppinion.
> 
> It's clear violation of internet standards that say "such packets MUST NOT
> be forwarded", so what else can it be but "a bug"?

Well I can only agree with you here, but I will come back later.

[...]

>> DHCP carries interface IDs in UDP payloads.  They're called "Link-layer
>> address" 48bit, in Client Identifier, as an UDP payload in DHCPv6
>> Solicit, by the User Terminal.
> 
> That is not a protocol commonly routed across the internet, so it is of
> total irrelevance in this context (and, using MAC address as client
> identifier is strictly optional in DHCPv6, unlike IPv4)

So, you seem to mean that the use of the IID in the payload of a DHCP 
Solicit is not necessarily mandatory.

So it is not because I dont set the right Client Identifier in the DHCP 
Solicit that I dont get replies.

>> If the Server is in the operator's network, then one could think that
>> there is no more tracking danger than without DHCP.  But if the Server
>> and intermediary routers are outside the operator's network then there's
>> real risk of tracking.
> 
> How many networks do you know that route DHCPv6 requests about arbitrary
> third party infrastructure?

Well, I think it will come.

The DHCPv6 requests could go very far - they're UDP.  I think the 
Solicit can go beyond link scope (site scope?), and can also be 
addressed to a GUA.

> Less thinking, more reading.  DHCPv6 initially was specified based on
> DUIDs, and *not* MAC addresses.

Initially yes.

Alex

> 
> Gert Doering
>          -- NetMaster
>