Date:Mon, 18 Mar 2002 09:36:09 -0800
From:Bob Hinden [EMAIL PROTECTED]
Message-ID: [EMAIL PROTECTED]
| As much as I would like to see it otherwise, except for link (and maybe
| subnet) scoped multicast, I don't think we can assume multicast across a
| wider
Thomas,
1) Do people think that the existance of multicast within a site
(e.g., home, enterprise, etc.) can be assumed?
Note: I'm assuming link-local multicast will always work, the
question is whether wider scope (beyond neighboring routers)
multicast is available.
As much as
Title: RE: Deployability/Useability of Multicast [was: Re: PPP and Global Addresses ]
DHCPv6 only requires link-local multicast not site or global. From draft-ietf-dhc-dhcpv6-23:
7.1. Multicast Addresses
DHCP makes use of the following multicast addresses:
All_DHCP_Agents (FF02::1:2
Toshi-san,
Jim-san,
Thank you for your thoughtful opinions and advices. Again, I
believe this
thank you for your kind words.
ISP-to-Customer site-configuration is one of the most
urgent issues for
the WG, because no non-technical customers can join to IPv6
world unless we
provide
Ole,
I agree but I believe we can have a stateless and stateful solution
for different needs.
I think prefix delegation is inherently stateful. both DHCP PD and
ICMP PD are as such stateful protocols. that is, the delegating party
has to keep track of which prefixes are assigned where.
The level 1 compliance stuff is only simple in the case of a single
subnet. As soon as you have routers between the client and the
server, you have to do something to get the packets across the router.
I suppose we could have a discussion about whether magic unicast
addresses (which are really a
On Wed, 13 Mar 2002 13:23:39 -0500,
Rob Austein [EMAIL PROTECTED] said:
The level 1 compliance stuff is only simple in the case of a single
subnet. As soon as you have routers between the client and the
server, you have to do something to get the packets across the router.
Correct, and
Jim-san,
Thank you for your thoughtful opinions and advices. Again, I believe this
ISP-to-Customer site-configuration is one of the most urgent issues for
the WG, because no non-technical customers can join to IPv6 world unless we
provide any solution. This positive discussion will help us to
At 02:28 PM 3/7/2002 -0500, Rob Austein wrote:
1) The portions of DHCP that are required for post-addr-conf (sorry,
don't have a better name for this) are pretty minimal, and I'm
pretty sure that one can write conforming DHCP clients and servers
that only implement that part of the
The DHCPv6 spec defines a well-known site-scoped multicast address
that all DHCP servers listen on. Assuming a DHCP client has
an address of sufficient scope to which a DHCP server can reply,
the client can send an Information-Request message to that
multicast address to contact a DHCp server
Title: RE: PPP and Global Addresses
Thanks for the pointer, the slides are very interesting indeed...
I totally agree this is an important topic which should be discussed on this list. IMO, the three mechanisms have different approaches and assets (for instance, APD takes routing
Jim and all,
[...]
To achieve zero-configuration for ISP-to-Customer (PE-to-CPE)
prefix delegation, we had better have a global consensus about
which mechanism is minimally required for this purpose. Many
mechnisms for the same purpose will make it difficult to achieve
zero-configuration
Jim and all,
Differentiate the need for an Intranet vs an Internet.
Most DHC is deployed on Intranets. Ralph has responded to you and I
concur.
I guess we are talking about what is the best for ISP-to-Customer
(PE-to-CPE) prefix delegation, correct?
Then, there seems three proposals shown
Toshi-san,
Jim and all,
Differentiate the need for an Intranet vs an Internet.
Most DHC is deployed on Intranets. Ralph has responded to you and I
concur.
I guess we are talking about what is the best for ISP-to-Customer
(PE-to-CPE) prefix delegation, correct?
I believe solving
Francis,
In your previous mail you wrote:
But I don't really care about your opinion or others on
what should be
used or not used from the work we do in the IETF.
= I disagree: the resources of IETF are not infinite so waste is
a common concern.
We are talking about use not a
just as note for those that don't know dhcpv6 supports multicast too.
/jim
-Original Message-
From: Rob Austein [mailto:[EMAIL PROTECTED]]
Sent: Saturday, March 09, 2002 12:34 PM
To: [EMAIL PROTECTED]
Subject: Re: PPP and Global Addresses
At Sat, 9 Mar 2002 03:27:36 +0100
At Sat, 9 Mar 2002 03:27:36 +0100, Hesham Soliman wrote:
I don't have a great knowledge of DHCP but wouldn't this simply
message still require a relay agent?
That's one way to do it. Another would be clever use of multicast.
[Replies intentionally redirected to mailing list by author -- I get
way too much mail and really don't need duplicate copies from a list
that generates as much traffic as this one. Thanks.]
At Thu, 07 Mar 2002 12:26:09 -0500, Ralph Droms wrote:
I do disagree with the DT conclusion - as do
1) The portions of DHCP that are required for post-addr-conf (sorry,
don't have a better name for this) are pretty minimal, and I'm
pretty sure that one can write conforming DHCP clients
and servers
that only implement that part of the DHCP spec (more
precisely, the
In your previous mail you wrote:
Catching up on old mail I saved for DHCPv6. Let me just start with your
view is wrong.
= do you argue we need dynamic address allocation for IPv6?
I don't see the point of arguing with you. DHCPv6 will be deployed and
widely used.
= Jim, you
In your previous mail you wrote:
But I don't really care about your opinion or others on what should be
used or not used from the work we do in the IETF.
= I disagree: the resources of IETF are not infinite so waste is
a common concern.
What I care about is if you find a technical
Francis,
I have to jump in here - DHCPv6 is *not* just for dynamic address
allocation. Have those who are claiming that DHCPv6 will not be used
actually read the spec? It will be used for other configuration
parameters, as described in draft-droms-dnsconfig-dhcpv6-01.txt Arguments
that
With all due respect, I've read draft-prigent-dhcpv6-threats-00.txt. The
authors based this doc on an old draft of the DHCPv6, which they did not
understand very well.
- Ralph
At 02:39 PM 3/7/2002 +0100, Francis Dupont wrote:
In your previous mail you wrote:
But I don't really care
In your previous mail you wrote:
I have to jump in here - DHCPv6 is *not* just for dynamic address
allocation. Have those who are claiming that DHCPv6 will not be used
actually read the spec?
= I read the spec (one of the statements I used in a message to Jim
is from the
At 05:52 PM 3/7/2002 +0100, Francis Dupont wrote:
In your previous mail you wrote:
It will be used for other configuration
parameters, as described in
draft-droms-dnsconfig-dhcpv6-01.txt Arguments
that DHCPv6 has no utility because of stateless address autoconfiguration
are
: [EMAIL PROTECTED]
Subject: Re: PPP and Global Addresses
In your previous mail you wrote:
Mr. Dupont, DHCP originally started with allowing dynamic
IP address allocation. A secondary benefit of utility is in
network operations, it is impossible to manually assign IP
address
Pekka,
I also think DHCP and what not should not be used for router
configuration.
In the military I had a seargent tell me once opinions are like rectums
every one has one.
Thats all this is and your entitled to it.
But I don't really care about your opinion or others on what should be
12:25 PM
To: [EMAIL PROTECTED]
Subject: Re: PPP and Global Addresses
IMO, there are two separate scenarios in which some sort of prefix
delegation needs to be used:
1) A router establishes a point-to-point connection to a provider.
2) A new router is connected to a site's backbone link
: RE: PPP and Global Addresses
I have to ask - because I've never heard it raised as an issue outside
this mailing list - has anyone ever heard a *network administrator* or
other customer of IETF protocols say that configuring
routers with what
is titled a host specific protocol will create
In your previous mail you wrote:
A point was made at the interim meeting (by Tony Hain, I believe, although
I trust my memory less now) that it would be prudent of us to ensure that
the ISP connection procedure for single hosts vs. routers is identical.
Otherwise, providers will be
PROTECTED]
Cc : [EMAIL PROTECTED]
Objet : Re: PPP and Global Addresses
PD (Automatic Prefix Delegation) doesn't specify any means to set the
prefix pool the routers rely on for delegation, apart from a manual
setting.
Could you kindly explain what you mean by this sentence in another way? I
did't get
PS: the conclusion is we should specify/finish/implement/deploy router
mechanisms ASAP, at least before IPv6 dialups become common.
I absolutely agree. My current conclusion among exsitsing choices is:
- APD for a customer whose CPE is a L3 Router
- RA(+RFC2462, 3041, etc.) for a customer
Title: RE: PPP and Global Addresses
PD (Automatic Prefix Delegation) doesn't specify any means to set the prefix pool the routers rely on for delegation, apart from a manual setting.
The DHCPv6 option could be used in this aim, and would be therefore complementary to PD...
Moreover PD also
Title: RE: PPP and Global Addresses
I did not attend the IPng meeting in May 2001 in Redmond and in the minutes I do not see the reasons why DHCPv6 protocol is not appropriate for router prefix delegation.
For several contexts, I think that we need some protocols or some extensions to make
On Fri, 15 Feb 2002, NOISETTE Yoann FTRD/DMI/CAE wrote:
PD (Automatic Prefix Delegation) doesn't specify any means to set the
prefix pool the routers rely on for delegation, apart from a manual setting.
The DHCPv6 option could be used in this aim, and would be therefore
complementary to PD...
I have to ask - because I've never heard it raised as an issue outside
this mailing list - has anyone ever heard a *network administrator* or
other customer of IETF protocols say that configuring routers with what
is titled a host specific protocol will create more confusion than it
is worth.
If
In your previous mail you wrote:
nope, not enough for a accounting, service termination etc.
= I know this argument but management is not allocation.
Regards
[EMAIL PROTECTED]
IETF IPng Working Group Mailing List
IPng
On Fri, 15 Feb 2002, Tony Hain wrote:
Ole Troan wrote:
...
would you be happier if we renamed it to SNCP (Simple Node
Configuration Protocol)? :-)
Actually, yes. Routers are not hosts, so configuring routers with what
is titled a host specific protocol will create more confusion than
IMO, there are two separate scenarios in which some sort of prefix
delegation needs to be used:
1) A router establishes a point-to-point connection to a provider.
2) A new router is connected to a site's backbone link.
From Brian Haberman's comments at the interim meeting, it seemed that APD
Title: RE: PPP and Global Addresses
PD (Automatic Prefix Delegation) doesn't specify any means to set the prefix pool the routers rely on for delegation, apart from a manual setting.
The DHCPv6 option could be used in this aim, and would be therefore complementary to PD...
Moreover PD also
actually delegate and route the
delegated prefix, and DHCP is not necessarily so.
Yoann
-Message d'origine-
De : Yamasaki Toshi [mailto:[EMAIL PROTECTED]]
Envoye : jeudi 14 fevrier 2002 04:52
A : Lilian Fernandes; [EMAIL PROTECTED]
Cc : [EMAIL PROTECTED]
Objet : Re: PPP and Global
Title: RE: PPP and Global Addresses
The draft-haberman-ipngwg-auto-prefix-01.txt says :
4.4 Prefix Delegation
After the request is verified to be acceptable, the Delegating
Router allocates the requested prefix size from its pool of
available addresses. The creation and management
In your previous mail you wrote:
... The creation and management of that pool is
beyond the scope of this document, but it can be supposed that
minimalistically a Delegating Router will be statically configured
with a fixed pool.
What I meant is that the pool used
In your previous mail you wrote:
Not even when the M and O bits in the Router Advertisement message are
configured to indicate a combination of stateless and stateful
autoconfiguration such that DHCPv6 provides options (e.g., SIP server
option) but not an actual address?
= look
In your previous mail you wrote:
Mr. Dupont, DHCP originally started with allowing dynamic
IP address allocation. A secondary benefit of utility is in
network operations, it is impossible to manually assign IP
address to 100's of hosts let alone 1,000,000's that IPv6
would allow.
nope, not enough for a accounting, service termination etc.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]On Behalf Of Francis Dupont
Sent: Friday, February 15, 2002 8:55 AM
To: Dr. Subrata Goswami
Cc: [EMAIL PROTECTED]
Subject: Re: PPP and Global Addresses
In your previous mail you wrote:
... The creation and management of that pool is
beyond the scope of this document, but it can be supposed that
minimalistically a Delegating Router will be statically configured
with a fixed pool.
What I meant is that the
Ole Troan wrote:
...
would you be happier if we renamed it to SNCP (Simple Node
Configuration Protocol)? :-)
Actually, yes. Routers are not hosts, so configuring routers with what
is titled a host specific protocol will create more confusion than it is
worth.
for me this boils down to
In your previous mail you wrote:
http://playground.sun.com/pub/ipng/html/minutes/ipng-meeting-may2001.txt
(search for Dialup Architecture, note this is my last attempt to find
an usefulness to DHCPv6 :-).
we're just about to suggest a couple of new DHCPv6 options for this
In your previous mail you wrote:
http://playground.sun.com/pub/ipng/html/minutes/ipng-meeting-may2001.txt
(search for Dialup Architecture, note this is my last attempt to find
an usefulness to DHCPv6 :-).
we're just about to suggest a couple of new DHCPv6 options for
In your previous mail you wrote:
I have a question about RFC2472 - IP Version 6 over PPP. The RFC talks
about the negotiation of the Interface-Identifier and specifies the
Interface-Identifier Configuration Option. In this case the upper 64 bits
are just fe80::
= yes, IPv6CP
In your previous mail you wrote:
Since PPP does not support multicast,
= I disagree:
x% uname -sr
FreeBSD 4.5-RELEASE
x% ifconfig ppp0
ppp0: flags=8010POINTOPOINT,MULTICAST mtu 1500
^
Is this what most implementations do?
= they implement
In your previous mail you wrote:
I have a question about RFC2472 - IP Version 6 over PPP. The RFC talks
about the negotiation of the Interface-Identifier and specifies the
Interface-Identifier Configuration Option. In this case the upper 64 bits
are just fe80::
= yes,
PD (Automatic Prefix Delegation) is one of the choices.
draft-haberman-ipngwg-auto-prefix-01.txt
http://search.ietf.org/internet-drafts/draft-haberman-ipngwg-auto-prefix-01.
txt
To solve all the configuration issues with DHCPv6 might be another good
choice, but I guess PD be a minimum
54 matches
Mail list logo
