---
--- IETF DRINKS WG, protocol design team minutes
--- May 21 2009
--- Attendees
Richard Shockey
Debbie Guyton
Ken Cartwright
Manjul Maharishi
Spencer Dawkins
Sumanth Channabasappa
Jean-Francois Mule
several dogs barking
--- Agenda
1/ Introductions and goals for 1st I-D
2/ Use cases and protocol requirements
3/ protocol I-D
+ scope
+ outline discussion
+ transport requirements
+ data model discussion
4/ Action item summary and next steps.
--- Notes
Minutes reported by Debbie and Jean-Francois.
Call moderated by Jean-Francois.
Any comments on these notes should be sent to drinks at ietf.org
We will have a weekly call, start with one hour and potentially extend
the call to two hours as we have more technical debates and issues to
close.
1/ Introductions of design team members and goals for 1st I-D
Our main goal is to produce a protocol that meets the requirements and
use cases discussed in DRINKS. Our immediate objective is to have a
first Internet-Draft by the end of June 2009 (before the cut-off date
for the next IETF).
Rich indicated that we should continue the work on use cases and
requirements in parallel.
Rich stated that there is consensus on a SOAP/XML protocol and there
is a need to sync with the Application Area folks, especially in the
security arena. This is something to keep in mind as progress is
made.
We discussed the outline for the general approach for developing a
protocol document:
- protocol capabilities
- transport requirements
- data model
- security
We then did a roundtable to get everyone's input on the goals and the
approach to make progress.
There was general agreement on the approach.
The protocol should provide options to extensibility as some people
may want to add elements to the data model later on (comment made by
several folks).
Debbie indicated that in XML-based protocols, it should be easier to
add extensibility. Ken used EPP and NETCONF as examples of IETF
protocols that lay out the approach to extensibility well and that
could be followed.
Richard thought that it would be useful to consult the app area AD
Lisa Dusseault for input, particularly on security requirements, and
perhaps have some discussions with the application ADs to get their
input.
# Action Item (AI): Jean-Francois before 5/28
# Jean-Francois to ping Lisa to get her on a call with this group
# (URL to charter, goals & requirements, minutes of the last meeting
# to give her the background).
2/ Use cases and protocol requirements
Sumanth suggested that we keep the use cases and requirements doc in
sync and volunteered to do so. Everyone agreed that would be good.
# AI: Sumanth on a regular basis by sending mails to IETF drinks
# list with deltas and proposed changes
# Sumanth to keep the use case and requirement documents in sync in
# case use cases need to be improved.
3/ Protocol I-D
Rich summarized the current wg consensus to use SOAP/XML (in both BoF
meetings) and each time we met to discuss that. Rich re-iterated some
of the comments he made in previous wg about the lack of EPP
implementation for SP for things like DRINKS want to do.
Ken added that the protocols like EPP and NETCONF have separated transport
from data model/structure definitions of the protocol. Ken
recommended we follow the same approach so it gives implementers the
option to implement another transport protocol to define new
transport. There was general agreement to follow this model.
- Scope of this protocol
Jean-Francois asked the question of the scope of the protocol.
Ken put the emphasis on the data model: it should meet both the needs
of the registry-to-registry (provisioning of the registry) and
registry to cache server.
We had unanimous consensus to do the client to registry protocol
definition first.
first - client to registry (front-end of provisioning the registry)
later - registry to server (which includes registry-to-registry)
- Document Outline
Ken volunteered to provide a first proposal for our next call based on
existing work done on provisioning and config protocols:
- RFC 4741 and RFC 4743 define NETCONF, and
- RFC 4930 and RFC 4934, EPP and EPP over TCP
We agreed to get a draft outline out before our next call based on
these examples. This could help speed up things.
Spencer commented that it would be good to keep the outline small for
now (agreed).
# AI: Ken before 5/27 COB
# Ken to send a first proposal to the design team via email so we can expedite
- Transport requirements
# AI: Jean-Francois to send input before 5/27 COB based on previous IETF DRINKS
# presentation done in Dublin to the list.
# AI: Debbie agreed to review her presentation and list required
# features of the protocol for design team discussion.
- Data model discussion
We had consensus that the data model should belong in the protocol
document, at least for now.
Manjul volunteered to provide an initial proposal for bashing on the next call.
Rich and Sumanth agreed that it would be better to have someone look at it.
Rich insisted that we should refrain from adding to much for it.
Jean-Francois agreed, we should keep things simple rather than try and
be exhaustive.
# AI: Manjul to propose a basic data model for team's review on the
# next call
Rich indicated we should take into account the WG comments about not
being too PSTN centric in the data model, a criticism that has been
often voiced given the use of PSTN and Routing Numbers in the use
cases.
- action item summary and next steps.
Jean-Francois reviewed the action item list (see above, denoted by
AI:).
We concluded the call.
# AI: Jean-Francois to consolidate his notes with Debbie and send
# them to the DRINKS WG list
Next call is Thursday May 28, 1pm MT, 3pm ET, and 9pm CET.
> end.
Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.