[IPv6]Re: I-D Action: draft-ietf-6man-rfc8504-bis-01.txt

[Jeremy Duncan <[e-mail]>]

Tim-

I’d like to help write that draft if it’s possible. We have a number of erratas that are tagged waiting document update that should be done as well.


-Jeremy

From: Timothy Winters <[e-mail]>
Sent: Thursday, July 10, 2025 1:01 PM
To: Jeremy Duncan <[e-mail]>
Cc: Ole Troan <[e-mail]>; Brian E Carpenter <[e-mail]>; 6man WG <[e-mail]>
Subject: Re: [IPv6]Re: I-D Action: draft-ietf-6man-rfc8504-bis-01.txt

[EXTERNAL] Verify links and attachments with sender.
Hi Jeremy,

I would also add RFC 7772 which recommends configuring Routers to respond with unicast RAs and send unsolicited RAs infuently.

If a host doesn't transmit an RS, it may wait a longer time to get any information about the network.

I'd be willing to see if we can write a short draft that updates this section in Neighbor Discovery.  I do think that IPv6 General purpose Host should always send a RS.

~Tim

On Thu, Jul 10, 2025 at 8:22 AM Jeremy Duncan <jduncan=[e-mail]<mailto:[e-mail]>> wrote:
Ole-

So these networks will not use unsolicited RAs at all? In the case of tunneled DSL it makes sense as the AN can't just send out tunneled RAs to just anyone. This is where it would make sense to require that for that use case and make it mandatory for tunnel DSL as it stipulates in RFC 6788.

If the consensus is to require RS in NDP, why can't we update 4861 to state as much if that was the original intent? This would alleviate a lot of confusion.

-Jeremy


-----Original Message-----
From: Ole Troan <[e-mail]<mailto:[e-mail]>>
Sent: Thursday, July 10, 2025 2:43 AM
To: Brian E Carpenter <[e-mail]<mailto:[e-mail]>>
Cc: Jeremy Duncan <[e-mail]<mailto:[e-mail]>>; 6man WG <[e-mail]<mailto:[e-mail]>>
Subject: Re: [IPv6]Re: I-D Action: draft-ietf-6man-rfc8504-bis-01.txt

[EXTERNAL] Verify links and attachments with sender.

>> Brian-
>> Yeah, that makes sense and would seem to be how that gets implemented. So in the bis, should we just say RS as a SHOULD instead of a MUST? And make listening and acting upon RAs a MUST? That's how I would like to see it at least.
>
> Either that, or explain why the SHOULD has been upgraded to a MUST.

RFC6788 and RFC7559 likely.
Certain networks use RS to build state and will not let hosts in that do not use RS.

Ole.


--------------------------------------------------------------------
IETF IPv6 working group mailing list
[e-mail]<mailto:[e-mail]>
List Info: https://mailman3.ietf.org/mailman3/lists/[e-mail]/
--------------------------------------------------------------------