[CDNi] Preparing draft-ietf-cdni-problem-statement for WG Last Call

[Ben Niven-Jenkins <[e-mail]>]

Colleagues,

At the last IETF I said I would try to prepare a version of the CDNI problem Statement (draft-ietf-cdni-problem-statement) that is ready for WG Last Call before IETF82 in Taipei.

As part of that I have reviewed the complete draft and made some minor editorial clean-ups to my working copy. However, there were a number of other areas where I think changes need to be made to the document that I think warrant wider WG review/consensus and I have listed them below along with the changes I am proposing.

Could you please provide your feedback on the areas/proposed text I have highlighted below as well as any other areas of the document that you think require additional work/cleanup before you consider the document is ready for a WG Last Call.

1) Section 2: Update the last paragraph to reference the WG draft for CDNI use cases and remove reference to draft-watson-cdni-use-cases, i.e.

OLD:
   Use cases for CDN Interconnection are further discussed in
   [I-D.bertrand-cdni-use-cases] (which contains a merged set of use
   cases previously presented in [I-D.watson-cdni-use-cases] and
   [I-D.bertrand-cdni-use-cases-00]).
NEW:
   Use cases for CDN Interconnection are further discussed in
   [I-D.ietf-cdni-use-cases].


2) Section 4 contains an introduction and 2 sub-sections (4.1 & 4.2) that are each just a couple of sentences, so I propose merging them into a single section and changing the first two sentences of section 4 appropriately, i.e.

OLD:
   This section expands on how CDNI interfaces can reuse and leverage
   existing protocols.  First the "reuse instead of reinvent" design
   principle is restated, then each inetrface is discussed individually
   with example candidate protocols that can be considered for reuse or
   leverage. 
NEW:
   This section expands on how CDNI interfaces can reuse and leverage
   existing protocols before describing each CDNI interface individually
   and highlighting example candidate protocols that could considered for
   reuse or leveraging to implement the CDNI interfaces.  


3) Section 5 is currently title "Gap Analysis of relevant Standardisation and Research Activities" but following the agreement at the last IETF to move the generic descriptions of other activities into an appendix Section 5 no longer references any research activities so I propose changing the title to" Gap Analysis of relevant Standardisation activities".

4) After the general introduction in Section 5 there are only sections that outline other research activities for Content Acquisition (section 5.1) and CDNI Metadata (Section 5.2) but there are not sections for the other CDNI interfaces (Control, Logging, Request Routing).

The reason for the lack of text is IMO because there is no current or past standardisation activity addressing those areas but the document currently does not mention that and therefore gives the impression of being incomplete.

I therefore propose making the following change to the last sentence in Section 5:

OLD:
   The following sections will summarize the existing work of the
   standard bodies above against the CDNI problem space.
NEW:
   The following sections will summarize the existing work of the
   standard bodies above against the CDNI problem space. Section 5.1
   summarises existing interfaces that could be leveraged for content
   acquisition between CDNs and Section 5.2 summarises existing metadata 
   specifications that may be applicable to CDNI. To date there is no
   existing standardisation activities in the areas of the remaining
   CDNI interfaces (CDNI Request Routing, CDNI Control and CDNI Logging).

5) I also propose using the following text for the third bullet in Section 5 unless someone can suggest more appropriate text to summarise in a sentence or two any CDN metadata related work in other standards bodies.

OLD:
   o  <TODO: Add a sentence on ITU>
NEW:
   o  CableLabs, SNIA and ITU have defined (or are working on)
      definitions for content related metadata definitions and specification
      for its distribution. However, they do not include metadata specific
      to the distribution of content within a CDN or between interconnected
      CDNs.

6) Section 10.1 lists a number of "normative references", if we adopt the proposed changes above it will end up containing "normative" references to:
   - draft-ietf-cdni-use-cases
   - draft-bertrand-cdni-experiements
   - draft-jenkins-cdni-names
   - RFC2119

After reading through the text of the Problem Statement again, I am not convinced that the Problem Statement references any of those documents normatively as the 3 CDNI related drafts are referenced in a context where they are for further information and the Problem Statement does not make use of any of the capitalised terms defined in RFC2119.

Therefore, unless anyone objects I will move the references for the three CDNI related drafts to the "Informative References" section and remove the reference for RFC2119.


Thanks
Ben