Directory Assistance Problem and High-Level Design Proposal
draft-ietf-trill-directory-framework-07

[Ballot discuss]
Ought to be quick:

s7 mentions "trusted directory services" for the first time.  Is the assumption there that the repository is trusted and what exactly do you mean by that or is it just a "dedicated" repository?

[Ballot comment]
There's a whole bunch of security considerations that affect repositories; many captured by LDAP.  But, I guess I'll save all those for the actual protocol ;)

I have to admit I'm with Joel too.

[Ballot comment]
A further WG last call explicitly pointing to the IPR claim has been issued and the completion of that addresses any issue I can have claimed was worthy of a Discuss. Personally, I am surprised that the WG is happy to go ahead in the face of an IPR disclosure where the license terms are not clear, but if they are OK with that then I am not about to demand that they stop implementing.

[Ballot comment]
Updated...

Given some additional discussions I have had on this document, I am strongly leaning more towards Joel J.'s view of this document.  It is not a framework, but it is also not a strong problem statement document.  I am unclear what purpose this document will serve and I am not sure how it could be changed to make it worth publishing.

I am curious about the sub-sections in section 5.  There is discussion of how RBridges can learn the mappings via either a Push or Pull model (and the corresponding requirements) and what information should be in the directory.  Was there discussion of adding a section discussing how the information gets placed in the directory (and the corresponding requirements)?  Something to consider, but is more a curiosity on my part and definitely non-blocking.

[Ballot comment]

- Why is there no definition of directory service in
section 2? For this reader that term means an X.500 or LDAP
like service, but I'm not at all clear if that's what's
envisaged.

- If an LDAP-like service is what is envisaged then I don't
see how a push model can work. Perhaps if you delve deep
enough into X.500 history there is a matching model with
DSA replication though, not sure. That'd be very complex
though.

- If what's envisaged is for the TRILL WG to develop an
entirely new directory service, then I think that should
be seriously discussed with the applications area. I'm
sure our apps ADs would take a keen interest were that
to be the case.

- Section 6 should IMO note the possibility that it may not
be possible to follow this recommendation, that is, failure
is a possible outcome.

- The security considerations should note that any
directory service would be a fine target to attack and
could constitute a single point of failure. The current
text all seems to be saying how good all this would be,
which I don't think is right.

- Managing access control in directories is always hard and
that should be noted here.

[Ballot comment]

- Why is there no definition of directory service in
section 2? For this reader that term means an X.500 or LDAP
like service, but I'm not at all clear if that's what's
envisaged.

- If an LDAP-like service is what is envisaged then I don't
see how a push model can work. Perhaps if you delve deep
enough into X.500 history there is a matching model with
DSA replication though, not sure. That'd be very complex
though.

- If what's envisaged is for the TRILL WG to develop an
entirely new directory service, then I think that should
be seriously discussed with the applications area. I'm
sure our apps ADs would take a keen interest were that
to be the case.

- Section 6 should IMO note the possibility that it may not
be possible to follow this recommendation, that is, failure
is a possible outcome.

- The security considerations should note that any
directory service would be a fine target to attack and
- Why is there no definition of directory service in
section 2? For this reader that term means an X.500 or LDAP
like service, but I'm not at all clear if that's what's
envisaged.

- If an LDAP-like service is what is envisaged then I don't
see how a push model can work. Perhaps if you delve deep
enough into X.500 history there is a matching model with
DSA replication though, not sure. That'd be very complex
though.

- If what's envisaged is for the TRILL WG to develop an
entirely new directory service, then I think that should
be seriously discussed with the applications area. I'm
sure our apps ADs would take a keen interest were that
to be the case.

- Section 6 should IMO note the possibility that it may not
be possible to follow this recommendation, that is, failure
is a possible outcome.

- The security considerations should note that any
directory service would be a fine target to attack and
could constitute a single point of failure. The current
text all seems to be saying how good all this would be,
which I don't think is right.

- Managing access control in directories is always hard and
that should be noted here.

could constitute a single point of failure. The current
text all seems to be saying how good all this would be,
which I don't think is right.

- Managing access control in directories is always hard and
that should be noted here.

[Ballot comment]
IMHO this isn't a framework at all. it's a problem, requirements, and two operational models and some non-specific recommendations for more work.

a framework is the structure that holds up the specification.

I am unclear what actions the working group is planning on doing on the basis of section 6.

6. Recommendation

  TRILL should provide a directory assisted approach.  This document
  describes a basic framework for directory assistance to RBridge edge
  nodes. More detailed mechanisms will be described in a separate
  document or documents.

The document appears to suggest implementing both mechanisms that it proposes, which is great I guess but normally I'd expect a document like this from a working group to arrive at a consensus solution instead of two of them.

[Ballot discuss]
I am surprised that the WG is willing to request the publication of
documents where there is known IPR, but the license terms have not been
disclosed. There are multiple interpretations of this including the WG
not considering that the IPR applies, the WG not believing that the IPR
is enforceable, the WG not having the issue properly exposed by the
chairs. or the WG being happy to consider paying a license for their
implementations.  Given that one of the chairs is the IPR holder, the
issue needs to be treated with a high degree of caution if for no other
reason than to protect the chair against claims of abuse.

The Shepherd write-up points to a fumble during WG last call, but
suggests the issue was handled with an additional call for comments
within the WG. The write-up also points at a slide from IETF-86. I
found this at
http://www.ietf.org/proceedings/86/slides/slides-86-trill-10.pdf
The disclosure (which was initially made against the -05 version of the
individual I-D, and later against the -04 version of the WG I-D) has
never included licensing terms.

Given that "IETF working groups prefer technologies with no known IPR
claims or, for technologies with claims against them, an offer of
royalty-free licensing" [RFC3979] I would like to understand the WG
process that led to this document being sent for publication.

I am sceptical that I am raising an issue that meets the IESG's Discuss
criteria, but I would like to discuss the point (at least with the
responsible AD) to understand that IETF process was followed
sufficiently to warrant deviating from the general case described in
RFC 3979. Then I will be able to move to No Objection with a clear
conscience.

[Ballot comment]
Donald proposed the following text to clear my Discuss, and it works for me:

> I suspect it would be good to add two sentences to the paragraph at
> the beginning of Section 5. Something like:
>      "It is optional whether or not an RBridge uses Push Directory
> services, Pull Directory services, or some combination. It is optional
> whether a Directory offers Push services, Pull services, or both."

Thank you for the quick response!

Donald also responded to my comment about collecting the text that compares and contrasts the Push and Pull models into one section, by saying that the working group hasn't agreed on what advice to give on when to Push and when to Pull, and that this is more appropriate for more detailed protocol specifications anyway.

That makes sense to me. Perhaps it's worth pulling the partial description of advantages and disadvantages out of the framework doc now and including a more complete description as the working group consenses, but as before, this is not a blocking comment.

[Ballot discuss]
Very readable. Thanks for that.

My Discuss is one question long, so I expect it's easy to clear.

Section 5 describes two models of operation (Push and Pull):

5. Generic operation of Directory Assistance

  There are two different models for Directory assistance to edge
  RBridges: Push Model and Pull Model.  The Directory Information is
  described in Section 5.1 below while Section 5.2 discusses Push Model
  requirements and Section 5.3 Pull Model requirements.

I'm not seeing a statement about whether all RBridges (or at least all RBridges used with TRILL directories) are expected to support both models of operation. I'm reading 5.3 Pull Model and Requirements as saying "yes, at least some RBridges would support both", but I'm guessing (the text I'm looking at is).

  For RBridges with mapping entries being pushed from a directory
  server, they can be configured to use the Pull Model for targets
  which don't exist in the mapping data pushed.

So, I'm asking if you can say anything helpful about that. Either "the answer depends" or "we don't need to say that in the framework document" could be fine answers, but I wanted to ask.

[Ballot comment]
I am finding some text that compares/contrasts the two models of operation, but it is scattered across the description of both models. Perhaps it would be more useful to the reader if this text could be collected into one section (perhaps a 5.4 "When to use each model of operation").

I don't know that any new text is needed - I'm asking about moving text like this, found in 5.2 Push Model and Requirements,

  However, the Push Model usually will push more entries of MAC&Label
  <-> Egress RBridge mapping to an edge RBridges than needed.  Under
  the normal process of edge RBridge cache aging and unknown
  destination address flooding, rarely used mapping entries would have
  been removed.  But it can be difficult for Directory Servers to
  predict the communication patterns among applications within one Data
  Label.  Therefore, it is likely that the Directory Servers will push
  down all the MAC&Label entries if there are end stations in the Data
  Label attached to the edge RBridge. This is a disadvantage of the
  Push Model compared with the Pull Model described below.

closer to text like this, found in 5.3 Pull Model and Requirements,

  One advantage of the Pull Model is that edge RBridges can age out
  MAC&Label entries if they haven't been used for a certain configured
  period of time or a period of time provided by the Directory.
  Therefore, each edge RBridge will only keep the entries that are
  frequently used, so its mapping table size will be smaller.  Edge
  RBridges would query the Directory Server(s) for unknown MAC
  destination addresses in data frames or ARP/ND and cache the
  response.  When end stations attached to remote edge RBridges rarely
  communicate with the locally attached end stations, the corresponding
  MAC&VLAN entries would be aged out from the RBridge's cache.

As we say, not a blocking comment ...

[Ballot comment]
I have no problems with the publication of this document.  I am curious about the sub-sections in section 5.  There is discussion of how RBridges can learn the mappings via either a Push or Pull model (and the corresponding requirements) and what information should be in the directory.  Was there discussion of adding a section discussing how the information gets placed in the directory (and the corresponding requirements)?  Something to consider, but is more a curiosity on my part and definitely non-blocking.

[Ballot comment]
-- Section 10.1 --

  As an Informational document, this draft has no Normative References.

I don't believe the general sentiment behind this (that Informational documents don't have normative references) is true.  I think normative references are those that must be understood in order to understand the document at hand, and that applies to documents of any status.

I think it would be helpful if the authors should sort the references with that in mind.

That said, this is a non-blocking comment.  Please consider doing that sort, but there's no need to respond to this comment.

IESG/Authors/WG Chairs:

IANA has reviewed draft-ietf-trill-directory-framework-05, which is
currently in Last Call, and has the following comment from the IANA's
reviewer:

We understand that, upon approval of this document, there are no IANA
Actions that need completion.

If this assessment is not accurate, please respond as soon as possible.

The following Last Call announcement was sent out:

From: The IESG
To: IETF-Announce
CC:
Reply-To: ietf@ietf.org
Sender:
Subject: Last Call:  (TRILL (Transparent Interconnection of Lots of Links): Edge Directory Assistance Framework) to Informational RFC


The IESG has received a request from the Transparent Interconnection of
Lots of Links WG (trill) to consider the following document:
- 'TRILL (Transparent Interconnection of Lots of Links): Edge Directory
  Assistance Framework'
  as Informational RFC

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 2013-07-18. 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


  Edge TRILL (Transparent Interconnection of Lots of Links) switches
  currently learn the mapping between MAC addresses and their egress
  TRILL switch by observing the data packets they ingress or egress or
  by the TRILL ESADI (End Station Address Distribution Information)
  protocol. When an ingress TRILL switch receives a data frame whose
  destination address (MAC&VLAN) that switch does not know, the data
  frame is flooded within the frame's VLAN across the TRILL campus.

  This document describes the framework for using directory services to
  assist edge RBridges in reducing multi-destination frames,
  particularly unknown unicast frames flooding, and ARP/ND, thus
  improving TRILL network scalability.






The file can be obtained via
http://datatracker.ietf.org/doc/draft-ietf-trill-directory-framework/

IESG discussion can be tracked via
http://datatracker.ietf.org/doc/draft-ietf-trill-directory-framework/ballot/


The following IPR Declarations may be related to this I-D:

  http://datatracker.ietf.org/ipr/2007/

(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?

Informational as noted in the title page header. This is a broad
overview of the idea for TRILL Directory Assistance. The WG
consensus for this document indicates the TRILL WG intent to pursue
this technical course as a method for the reduction of
multi-destination traffic, but this document does not include
protocol specifications.

(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 contains a brief overview of difficulties cause by
multi-destination local network traffic, such as unicast frames
where the location of the destination is unknown or ARP/ND packets,
and a generic description of how to reduce such difficulties
through the use of push and/or pull directories. Data Centers are
frequently orchestrated by management software that maintains a
directory of the applications, Virtual Machines, IP and MAC
addresses in use and the switches to which they are attached. This
makes directory assistance to edge switches a reasonable strategy
for that case.

Working Group Summary:

The WG Last Call reviewers of this document include David Black,
Sam Aldrin, Erik Nordmark, and Yizhou Li. Some of their comments
suggested the removal of material not directly related to the main
topic of the document as well as other improvements. Those
suggestions were generally adopted. No parts of the draft were
particularly contentious.

Document Quality:

The document is of good quality. David Black's review was
particularly helpful and is very much appreciated.

Personnel:

Document Shepherd: Jon Hudson
Responsible Area Director: Ted Lemon

(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.

Several minor issues with clarity, grammar etc were found and corrected.
The Document showed up in good shape and all major issues has been
handled during earlier WG review. Overall the process went smoothly.

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

None.

(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.

No specific additional review required.

(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.

The adoption and approval of this document as informational
indicates WG approval of work to specify directory assistance
mechanisms as an optional way to reduce multi-destination traffic
in TRILL campuses. As a result, there are no specific concerns.

(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.

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

Yes, see IPR Disclosure 2007. This disclosure has been brought to
the attention of the WG on many occasions, although it was
initially omitted in the WG Last Call. It was mentioned in the call
for the draft to be made a WG document (in that case, the same
disclosure but against the earlier personal draft version), posted
to the WG mailing list with an additional call for comments after
the WG Last Call, and specifically presented at the TRILL WG
meeting in Orlando at IETF-86 with a dedicated slide.

(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 good consensus for this document. This is most likely
due to extensive reviews, the resolution all comments in those
reviews on the mailing list as well as multiple presentations and
discussion at WG meetings.

(10) Has anyone threatened an appeal or otherwise indicated extreme
discontent? If so, please summarize 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.

(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.

None found.

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

No formal review required.

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

Yes, although, as an Informational document, there are no Normative
References.

(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.

(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?

This document does not change the status of any existing RFC.

(17) Describe the Document Shepherd's review of the IANA
considerations section, especially with regard to its consistency with
the body of the document.

This document requires no IANA actions.

(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.

This document does not create any new IANA Registries.

(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.

This document uses no formal languages requiring such validation.
////////////////