Triggering DHCPv6 Reconfiguration from Relay Agents
draft-ietf-dhc-triggered-reconfigure-07

[Ballot discuss]
-- Section 6.6 --

  The relay agent assumes the
  request is successfully handled for a client if the corresponding
  Client Identifier Option does not appear in at least one Reconfigure-
  Reply message.

I'm not sure whether what this seems to say is what you mean to say.  Perhaps that means that it should be re-worded, either way.

What it says is that if the Client Identifier Option (I'll say "CIO") does not appear in *any* Reconfigure-Reply messages, then the relay agent assumes the request is successfully handled.  But if the CIO appears in at least one Reconfigure-Reply message, then the relay agent can't assume anything.

What I *think* you might mean is that if there is at least one Reconfigure-Reply message in which the CIO does not appear, then the relay agent assumes the request is successfully handled.  But if the CIO appears in *all* the Reconfigure-Reply message, then the relay agent can't assume anything.

Which do you mean?  Can you re-word this to make it clear?

[Ballot comment]
The following former blocking comments have been resolved.

-- Section 5 --

  The relay agent SHOULD use the same value that was sent
  to the DHCPv6 server when relaying messages from the client to the
  server, as in Section 20.1.1 of [RFC3315].

Suggestion:
NEW
  To defend against poor implementations that do not correctly evaluate
  equivalence, the relay agent SHOULD use the same value that was sent
  to the DHCPv6 server when relaying messages from the client to the
  server, as in Section 20.1.1 of [RFC3315].

-- Section 6.3 --

  The relay may receive Reconfigure encapsulated in Relay-Reply before
  Reconfigure-Reply.  The relay SHOULD NOT interpret it as if the
  Reconfigure-Request was successfully handled by the Server.  The
  relay SHOULD use Reconfigure-Reply, not the Reconfigure message, to
  determine if the request was successful.

Suggestion:
NEW
  relay SHOULD use Reconfigure-Reply, not the Reconfigure message, to
  determine if the request was successful (see the discussion in Section 4).

---------------------------------------------------------
The following non-blocking comments have been resolved.

-- Section 3 --

  [RFC6422] updates the DHCPv6 specification with a new feature to let
  a DHCPv6 relay agent communicate information towards a DHCPv6 client,
  and which is not available at the DHCPv6 server.

That last phrase, starting with "and", doesn't seem to fit.  What is it that's not available at the server?  The new feature?  Maybe it's the "and" that's confusing me, but I think this needs to be rephrased.

  The change of the configuration may result in an exchange of CoA
  (Change-of-Authorization, [RFC5176]) messages

What change of what configuration?  The use of "the" implies that there's something we should know about that you're referring to?  Do you just mean "A configuration change"?  Are we specifically talking about "A configuration change to the DAC server"?  Or can the configuration change also be in the DHCPv6 relay?  Or what?

  Note the change of the configuration in the DHCPv6 relay agent can be
  triggered by any other out-of-band mechanism.

You're not already talking about an out-of-band mechanism, to justify "other", as far as I can tell.  I think you mean "by some out-of-band mechanism," or "by an out-of-band mechanism," yes?  Are there in-band mechanisms as well, which aren't shown here?  The paragraph after the diagram implies that there are.

-- Section 6.2.1 --

  Servers MUST discard any received RECONFIGURE-REQUEST messages that
  meet any of the following conditions:

The other three discard instructions in Section 6.2 say "MUST silently discard".  Is there a reason this one doesn't have "silently", or was it just an oversight?

-- Section 6.3 --

  In case multiple servers are configured to the relay agent, several
  Reconfigure-Request messages are to be built.

Is there a connection between "multiple" and "several" here?  Is this meant to say this?:

NEW?
  When multiple servers are configured to the relay agent, multiple
  Reconfigure-Request messages are to be built, one per server.

Or is it trying to say something else?  If so, what?

-- Section 7 --

  The relay MUST be configured with required rate-limit
  parameters (i.e., the rate of Reconfigure-Request messages).

That's the only rate-limit parameter that's required: the rate of Reconfigure-Request messages?  If that's not the case, please fix what's inside the parentheses.  (Hint: I would just delete the parenthesized bit.)  The same goes for the subsequent paragraph.

-- Section 9 --

  In addition,
  DHCPv6 servers MAY be configured to reject relayed Reconfigure-
  Request messages or restrict relay chaining (see [RFC5007] for more
  discussion about the rationale of this recommended behavior).

This is NOT "recommended behavior": MAY is not a recommendation.  MAY means that what you're talking about is entirely optional.  If you want to recommend this, you need to use SHOULD, if that's appropriate, or simply skip the 2119 language and say it in English.

[Ballot comment]
I have no problems with the publication of this document, but I do have some non-blocking comments...

1. Section 6 : The first sentence in 6.2.1 ("Clients MUST silently discard any received RECONFIGURE-REQUEST message") is useless.  A client implementation won't know about this protocol that works between relays and servers.

2. Section 6.3 : I see "The relay agent MAY supply the updated configuration in the RSOO" and wonder why this is a MAY. If the updated configuration is not included, what is the DHCP server to do? The discussion in Section 6.5 does not seem to cover the situation from the server's perspective.

[Ballot comment]
1. Section 3

  [RFC6422] updates the DHCPv6 specification with a new feature to let
  a DHCPv6 relay agent communicate information towards a DHCPv6 client,
  and which is not available at the DHCPv6 server.

This doesn't match the RFC 6422 abstract, on the top of being confusing ("which is not available at the DHCPv6 server
" refers to the "information", and not "DHCPv6 client", I guess, but your sentence doesn't say that)

RFC 6422 abstract:
  DHCPv6 relay agents cannot communicate with DHCPv6 clients directly.
  However, in some cases, the relay agent possesses some information
  that would be useful to the DHCPv6 client.  This document describes a
  mechanism whereby the DHCPv6 relay agent can provide such information
  to the DHCPv6 server, which can, in turn, pass this information on to
  the DHCP client.

Maybe you want something such as:

  For cases where the DHCPv6 relay agent possesses some information
  that would be useful to the DHCPv6 client, [RFC6422] specifies a
  mechanism whereby the DHCPv6 relay agent can provide such information
  to the DHCPv6 server, which can, in turn, pass this information on to
  the DHCP client.

2. Question: how does the DHCP server know if the DHCP relays support this reconfigure feature?
In other words, how does the DHCP server know if it can rely on this feature from his connected DHCP relays?

[Ballot comment]
6.4/6.5: "MUST be configurable"? Aside from this being completely unverifiable, I don't see any justification for such a requirement in this document.

6.5: "the server MAY use its content", "The server MAY use the content". Lowercase the MAYs. Better yet, use "might" or "can". These aren't protocol options.

[Ballot discuss]
s6.9: Because the Reconfigure option requires that the client authenticate it I'm happy (s15.11 of RFC 3315), but why wouldn't you also require that the rely agent & server support the authentication option as opposed to "A control policy based on the content of received Relay Identifier Option MAY be enforced by the DHCPv6 server"?

[Ballot comment]

s9:  Points to s21.1 of RFC 3315.  That section points to RFC 2401.  That RFCs been obsoleted.  Is there any chance that DHCPv6 is going to be looking at updating that?


s6.5: Worth adding RFC 5007 after the RFC 3315 for status codes because that's where Not Configured is defined.

s9: Worth adding a reference to RFC 6925 for the relay identifier option?  I had to go hunt that one down.

[Ballot discuss]
-- Section 5 --

  The relay agent SHOULD use the same value that was sent
  to the DHCPv6 server when relaying messages from the client to the
  server, as in Section 20.1.1 of [RFC3315].

Why SHOULD and not MUST?  Or, alternatively, why do you care at all?  What's the effect of using an equivalent, but different value?  How does an implementor know what the consequences are of obeying or of violating this SHOULD?

-- Section 6.3 --

  The relay may receive Reconfigure encapsulated in Relay-Reply before
  Reconfigure-Reply.  The relay SHOULD NOT interpret it as if the
  Reconfigure-Request was successfully handled by the Server.  The
  relay SHOULD use Reconfigure-Reply, not the Reconfigure message, to
  determine if the request was successful.

What effect does this have on the operation of the protocol?  What happens if a relay *does* use the Reconfigure message for this purpose?  How would an implementation evaluate the consequences here, to determine when it's OK to violate the SHOULD/SHOULD NOT?

-- Section 6.6 --

  The relay agent assumes the
  request is successfully handled for a client if the corresponding
  Client Identifier Option does not appear in at least one Reconfigure-
  Reply message.

I'm not sure whether what this seems to say is what you mean to say.  Perhaps that means that it should be re-worded, either way.

What it says is that if the Client Identifier Option (I'll say "CIO") does not appear in *any* Reconfigure-Reply messages, then the relay agent assumes the request is successfully handled.  But if the CIO appears in at least one Reconfigure-Reply message, then the relay agent can't assume anything.

What I *think* you might mean is that if there is at least one Reconfigure-Reply message in which the CIO does not appear, then the relay agent assumes the request is successfully handled.  But if the CIO appears in *all* the Reconfigure-Reply message, then the relay agent can't assume anything.

Which do you mean?  Can you re-word this to make it clear?

[Ballot comment]
Even though this set is non-blocking, I think some of these issues make the document less clear than it should be, and I strongly suggest that you consider these:

-- Section 3 --

  [RFC6422] updates the DHCPv6 specification with a new feature to let
  a DHCPv6 relay agent communicate information towards a DHCPv6 client,
  and which is not available at the DHCPv6 server.

That last phrase, starting with "and", doesn't seem to fit.  What is it that's not available at the server?  The new feature?  Maybe it's the "and" that's confusing me, but I think this needs to be rephrased.

  The change of the configuration may result in an exchange of CoA
  (Change-of-Authorization, [RFC5176]) messages

What change of what configuration?  The use of "the" implies that there's something we should know about that you're referring to?  Do you just mean "A configuration change"?  Are we specifically talking about "A configuration change to the DAC server"?  Or can the configuration change also be in the DHCPv6 relay?  Or what?

  Note the change of the configuration in the DHCPv6 relay agent can be
  triggered by any other out-of-band mechanism.

You're not already talking about an out-of-band mechanism, to justify "other", as far as I can tell.  I think you mean "by some out-of-band mechanism," or "by an out-of-band mechanism," yes?  Are there in-band mechanisms as well, which aren't shown here?  The paragraph after the diagram implies that there are.

-- Section 6.2.1 --

  Servers MUST discard any received RECONFIGURE-REQUEST messages that
  meet any of the following conditions:

The other three discard instructions in Section 6.2 say "MUST silently discard".  Is there a reason this one doesn't have "silently", or was it just an oversight?

-- Section 6.3 --

  In case multiple servers are configured to the relay agent, several
  Reconfigure-Request messages are to be built.

Is there a connection between "multiple" and "several" here?  Is this meant to say this?:

NEW?
  When multiple servers are configured to the relay agent, multiple
  Reconfigure-Request messages are to be built, one per server.

Or is it trying to say something else?  If so, what?

-- Section 7 --

  The relay MUST be configured with required rate-limit
  parameters (i.e., the rate of Reconfigure-Request messages).

That's the only rate-limit parameter that's required: the rate of Reconfigure-Request messages?  If that's not the case, please fix what's inside the parentheses.  (Hint: I would just delete the parenthesized bit.)  The same goes for the subsequent paragraph.

-- Section 9 --

  In addition,
  DHCPv6 servers MAY be configured to reject relayed Reconfigure-
  Request messages or restrict relay chaining (see [RFC5007] for more
  discussion about the rationale of this recommended behavior).

This is NOT "recommended behavior": MAY is not a recommendation.  MAY means that what you're talking about is entirely optional.  If you want to recommend this, you need to use SHOULD, if that's appropriate, or simply skip the 2119 language and say it in English.

[Ballot comment]
I have no objection to the publication of this document, but I do see a few issues that I would strongly suggest you resolve before handing the text to the RFC Editor.

---

Section 4 could probably be renamed since, I assume, this is no longer
a proposal, but a real protocol solution. Changes to the text as well.

---

In 6.4 I think the term "reject" is mis-applied because there is no
rejection involved.

---

Section 6.5 does not clearly explain how to build a rejection
Reconfigure-Reply message despite that section 9 says it does.

You should add a statement about how a rejection is conveyed and which
status code to use.

---

Section 6.6

  Depending on the status code enclosed in a received RECONFIGURE-REPLY
  message, the relay may decide to terminate the request or try a
  different corrected Reconfigure-Request.

Undoubtedly true, but which codes get which treatment?

---

Please decide whether the names of the new messages are in upper case
or not.

[Ballot comment]
minor nit:

section 4

  If the relay has no client  to reconfigured, it stops sending Reconfigure-Request messages.
-
  If the relay has no client to be reconfigured, it stops sending Reconfigure-Request messages.

or

  If the relay has no client to reconfigure, it stops sending Reconfigure-Request messages.

IESG/Authors/WG Chairs:

IANA has reviewed draft-ietf-dhc-triggered-reconfigure-05.  Authors should
review the comments and/or questions below.  Please report any
inaccuracies and respond to any questions as soon as possible.

We understand that, upon approval of this document, there are two actions which it must complete.

First in the DHCP Message types subregistry of the Dynamic Host Configuration Protocol for IPv6 (DHCPv6) registry located at:

http://www.iana.org/assignments/dhcpv6-parameters.xml

two new message types are to be registered as follows:

Value: [ TBD-at-registration ]
Description: RECONFIGURE-REQUEST
Reference: [ RFC-to-be ]

Value: [ TBD-at-registration ]
Description: RECONFIGURE-REPLY
Reference: [ RFC-to-be ]

Second, in the DHCPv6 Option Codes subregistry of the in Dynamic Host Configuration Protocol for IPv6 (DHCPv6) registry located at:

http://www.iana.org/assignments/dhcpv6-parameters.xml

a single new option code is to be registered as follows:

Value: [ TBD-at-registration ]
Description: OPTION_LINK_ADDRESS
Reference: [ RFC-to-be ]

The DHCPv6 Option Codes subregistry is managed via Expert Review and Specification Required as defined by RFC5226.

NOTE: We have initiated a request and sent this to the designated expert
for review.

We understand that these are the only actions required upon approval
of this document.

Note:  The actions requested in this document will not be completed
until the document has been approved for publication as an RFC.
This message is only to confirm what actions will be performed.

The following Last Call announcement was sent out:

From: The IESG
To: IETF-Announce:;
CC:
Bcc:
Reply-To: IETF Discussion List
Sender:
Subject: Last Call:  (Reconfigure Triggered by DHCPv6 Relay Agents) to Proposed Standard


The IESG has received a request from the Dynamic Host Configuration WG
(dhc) to consider the following document:
- 'Reconfigure Triggered by DHCPv6 Relay Agents'
  as Proposed Standard

The IESG plans to make a decision in the next few weeks, and solicits
final comments on this action. Please send substantive comments to the
ietf@ietf.org mailing lists by 2013-05-06. Exceptionally, comments may be
sent to iesg@ietf.org instead. In either case, please retain the
beginning of the Subject line to allow automated sorting.

Abstract


  This document defines new DHCPv6 messages: Reconfigure-Request and
  Reconfigure-Reply.  Reconfigure-Request message is sent by a DHCPv6
  relay agent to notify a DHCPv6 server about a configuration
  information change, so that the DHCPv6 server can send a Reconfigure
  message accordingly.  Reconfigure-Reply message is used by the server
  to acknowledge the receipt of Reconfigure-Request.

  This document updates RFC 3315 and RFC 6422.




The file can be obtained via
http://datatracker.ietf.org/doc/draft-ietf-dhc-triggered-reconfigure/

IESG discussion can be tracked via
http://datatracker.ietf.org/doc/draft-ietf-dhc-triggered-reconfigure/ballot/


No IPR declarations have been submitted directly on this I-D.

(1) What type of RFC is being requested (BCP, Proposed Standard,
    Internet Standard, Informational, Experimental, or Historic)? Why
    is this the proper type of RFC? Is this type of RFC indicated in
    the title page header?

Proposed Standard.  This documents an extension to the DHCPv6
protocol, so it only really makes sense as a PS.  The intended type is
indicated in the header.

(2) The IESG approval announcement includes a Document Announcement
    Write-Up. Please provide such a Document Announcement
    Write-Up. Recent examples can be found in the "Action"
    announcements for approved documents. The approval announcement
    contains the following sections:

Technical Summary:

This document defines new DHCPv6 messages: Reconfigure-Request and
Reconfigure-Reply. Reconfigure-Request message is sent by a DHCPv6
relay agent to notify a DHCPv6 server about a configuration
information change, so that the DHCPv6 server can send a Reconfigure
message accordingly. Reconfigure-Reply message is used by the server
to acknowledge the receipt of Reconfigure-Request.

Working Group Summary:

This document was proposed in the working group in August 2012 as an
extension to trigger client reconfiguration, via the DHCPv6 server,
when configuration changes are made on the relay agent. The work was
adopted as a WG item in September 2012 as there was good support and
interest for the work.

Document Quality:

I'm not aware of any existing implementations. This is something that
a number of folks and relay agent vendors have requested, and we
expect to see implementation in relay agents and servers.. I think
everyone who provided substantive review of the document is mentioned
in the acknowledgements section.

Personnel:

Bernie Volz is the document shepherd. Ted Lemon is the responsible AD.

(3) Briefly describe the review of this document that was performed by
    the Document Shepherd. If this version of the document is not
    ready for publication, please explain why the document is being
    forwarded to the IESG.

I read the document thoroughly, and submitted quite a few editorial
suggestions to the authors, which they implemented.  The differences
between the -02 and -05 versions of the document reflect this review.

(4) Does the document Shepherd have any concerns about the depth or
    breadth of the reviews that have been performed?

No, the document has had a great deal of careful review.

(5) Do portions of the document need review from a particular or from
    broader perspective, e.g., security, operational complexity, AAA,
    DNS, DHCP, XML, or internationalization? If so, describe the
    review that took place.

This is strictly a DHCP doc, and has had plenty of review from DHCP
experts.

(6) Describe any specific concerns or issues that the Document
    Shepherd has with this document that the Responsible Area Director
    and/or the IESG should be aware of? For example, perhaps he or she
    is uncomfortable with certain parts of the document, or has
    concerns whether there really is a need for it. In any event, if
    the WG has discussed those issues and has indicated that it still
    wishes to advance the document, detail those concerns here.

I think the document is good as written, and serves a useful purpose.


(7) Has each author confirmed that any and all appropriate IPR
    disclosures required for full conformance with the provisions of
    BCP 78 and BCP 79 have already been filed. If not, explain why?

Yes, the authors, Mohamed Boucadair and Xavier Pougnard, did confirm
that they are not aware of any IPR.

(8) Has an IPR disclosure been filed that references this document? If
    so, summarize any WG discussion and conclusion regarding the IPR
    disclosures.

No.

(9) How solid is the WG consensus behind this document? Does it
    represent the strong concurrence of a few individuals, with others
    being silent, or does the WG as a whole understand and agree with it?

There is a strong consensus behind this document and in particular from
very active WG participants (i.e. "DHCP experts").

(10) Has anyone threatened an appeal or otherwise indicated extreme
    discontent? If so, please summarise the areas of conflict in
    separate email messages to the Responsible Area Director. (It
    should be in a separate email because this questionnaire is
    publicly available.)

No, and nobody's indicated that they were against the WGLC or had
any issues with the document.

(11) Identify any ID nits the Document Shepherd has found in this
    document. (See http://www.ietf.org/tools/idnits/ and the
    Internet-Drafts Checklist). Boilerplate checks are not enough;
    this check needs to be thorough.

The document passes idnits with no errors and review using the
checklist did not turn up any issues.

(12) Describe how the document meets any required formal review
    criteria, such as the MIB Doctor, media type, and URI type
    reviews.

The document contains nothing like this.

(13) Have all references within this document been identified as
    either normative or informative?

Yes.

(14) Are there normative references to documents that are not ready
    for advancement or are otherwise in an unclear state? If such
    normative references exist, what is the plan for their
    completion?

No, all the normative references are to RFCs.

(15) Are there downward normative references references (see RFC
    3967)? If so, list these downward references to support the Area
    Director in the Last Call procedure.

No.

(16) Will publication of this document change the status of any
    existing RFCs? Are those RFCs listed on the title page header,
    listed in the abstract, and discussed in the introduction? If the
    RFCs are not listed in the Abstract and Introduction, explain
    why, and point to the part of the document where the relationship
    of this document to the other RFCs is discussed. If this
    information is not in the document, explain why the WG considers
    it unnecessary.

The draft (05) currently states that it updates RFC 3315 and RFC 6422.
This should be removed as this draft makes no changes to text in
either draft.

(17) Describe the Document Shepherd's review of the IANA
    considerations section, especially with regard to its consistency
    with the body of the document. Confirm that all protocol
    extensions that the document makes are associated with the
    appropriate reservations in IANA registries. Confirm that any
    referenced IANA registries have been clearly identified. Confirm
    that newly created IANA registries include a detailed
    specification of the initial contents for the registry, that
    allocations procedures for future registrations are defined, and
    a reasonable name for the new registry has been suggested (see
    RFC 5226).

I reviewed the IANA Considerations section and it is fine and clear;
two new DHCPv6 options are being requested (registry is clearly
identified).

(18) List any new IANA registries that require Expert Review for
    future allocations. Provide any public guidance that the IESG
    would find useful in selecting the IANA Experts for these new
    registries.

None.

(19) Describe reviews and automated checks performed by the Document
    Shepherd to validate sections of the document written in a formal
    language, such as XML code, BNF rules, MIB definitions, etc.

There are no such parts to the document.