[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[DNA] DAD in using DHCPv6

Hello all,

I've asked following question to DHC wg and IPv6 WG.
Then, they recommended to ask to this WG.

Could you give me your thought, please?
And I attached the emails that I've received in the WGs.

---------------------
In the situation in which DHCPv6 clinet transmits a Conform message for
its Global address,
should the DHCPv6 Client do DAD for GA?

When the DAD NS should be transmitted, if it is required to transmit a
DAD NS from DHCPv6 Client?
Should the DAD NS be transmitted after sending the Conform messge and/or
receiving the Reply? Or before sending the Conform message?

-----------------------

Thank you for your cooperation.

Best regards,
Hideshi


Hideshi:

There are really two parts to this:
1. Is DAD for the link-local done when an interface "reconnect" is
detected? This may be required before multicasting the Confirm.
2. Is DAD for done for the global addresses (AFTER a successful
indication from a DHCPv6 Confirm is received, since there's no point to
do it before as the client doesn't know yet whether the addresses are
invalid)?

I'm not sure that this is a DHCP question ... Probably should be
directed at the IPv6 WG?

- Bernie

-----Original Message-----
From: Hideshi.Enokihara@jp.yokogawa.com
[mailto:Hideshi.Enokihara@jp.yokogawa.com] 
Sent: Monday, August 20, 2007 10:26 PM
To: dhcwg@ietf.org
Cc: ipv6@ietf.org
Subject: Is a DNS NS required,when a DHCPv6 Client transmits a Conform
message? 

Hello all,

In the situation in which DHCPv6 clinet transmits a Conform message for
its Global address,
should the DHCPv6 Client do DAD?

When the DAD NS should be transmitted, if it is required to transmit a
DAD NS from DHCPv6 Client?
Should the DAD NS be transmitted after sending the Conform messge and
receiving the Reply? Or before sending the Conform message?

I'm waiting for your valuable replies.

Best regards,
Hideshi

--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------



At Mon, 20 Aug 2007 23:36:50 -0400,
"Bernie Volz (volz)" <volz@cisco.com> wrote:

> There are really two parts to this:
> 1. Is DAD for the link-local done when an interface "reconnect" is
> detected? This may be required before multicasting the Confirm.
> 2. Is DAD for done for the global addresses (AFTER a successful
> indication from a DHCPv6 Confirm is received, since there's no point to
> do it before as the client doesn't know yet whether the addresses are
> invalid)?
> 
> I'm not sure that this is a DHCP question ... Probably should be
> directed at the IPv6 WG?

I don't think it's a DHCPv6 question, either.  The real question is
whether a host should perform DAD when it detects a possible change of
attached link.  Referring to Section 18.1.2 of RFC3315, those case
would include:

   o  The host (client) reboots.
   o  The host (client) is physically connected to a wired connection.
   o  The host (client) returns from sleep mode.
   o  The host (client) using a wireless technology changes access points.

I think it's reasonable to perform DAD for at least some of these
cases, but as far as I know no specification explicitly requires that.

					JINMEI, Tatuya
					Communication Platform Lab.
					Corporate R&D Center, Toshiba Corp.
					jinmei@isl.rdc.toshiba.co.jp



> I'm not sure that this is a DHCP question ... Probably should be
> directed at the IPv6 WG?

Or perhaps DNA, which is grappling with the general issue of how to
deal (quickly) with reattachments to the same link, and what should
done in such cases.

Thomas