[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Draft minutes from Minneapolis
Very many thanks to Eric Gray, Dimitri Papadimitriou
and Deborah Brungard for taking minutes.
Please be gentle.
CCAMP Working Group.
58th IETF Minneapolis
Monday 10th November 2003 0900-11.30
Kireeti Kompella <firstname.lastname@example.org>
Adrian Farrel <email@example.com>
Agenda bashing (chairs)
Minute takers and blue sheets (chairs)
Minute takers: Dimitri Papadimitriou
Thanks to Ron Bonica for co-chairing
Thanks to Adrian Farrel for taking over as co-chair
Old charter (chairs)
Groundwork and critical work is completed, thus re-charter.
New charter (chairs)
- Multi-area/multi-as using tewg mpls requirements as input
- GMPLS ASON (input from ITU-T) look at the requirements and
then address them
- Short charter for 6 months and move forward
- Lots of interactions with the MPLS and IGP WG's
- Short term focus of the WG on the charter milestones
therefore new material at bottom of the agenda (thus take
it to the list if this is not covered during this meeting)
Drafts in 'final stages' (chairs)
- Routing drafts: comments received from IESG (cleared
- 2 drafts in IETF 4 weeks last call: LMP-SONET-SDH and
- 2 drafts pending: LMP-MIB and SDH/SONET-Control (with Bert
and Alex for review)
Work in progress (chairs)
- GMPLS UNI (overlay): needs one week WG last call
- GMPLS for G.709: look for WG last call but positive
comments needed to move forward so will ask via mailing
list if there is interest or not in progressing and if any
- Routing exclusion: new version is imminent (revision to be
published just after the meeting)
Summary of interactions with other WGs (chairs)
- TE WG: Multi-area/AS requirements (mpls only)
- MPLS WG: P2MP Requirements
- OSPF/IS-IS WG: GMPLS extensions completed now starting new
item on ASON Routing requirements (Design
- IPO WG: Framework document
ITU-T Liaison (Wesam Alanqar)
- Kireeti Kompella: noted that an ITU Recommendation can
have CR-LDP as a normative reference, that's ok, but for
future need to discuss among chairs/ADs (this will be done
- Alex Zinin: CR-LDP code-points liaison, for the time this
normative reference is ok but for the future will need to
clarify the liaison
- Kam Lam: ITU-T SG15/Q14 Feb04 interim meeting, in San Jose
or North Carolina (not decided yet) - invites participants
- Kam Lam: setup of a common FTP server / website
GMPLS MIBs (Tom Nadeau)
Tom Nadeau presented an update on the GMPLS MIB status.
- Three drafts - fairly stable, one more round of IESG
review for MPLS MIBs (these MIBs depend on the status
of the MPLS MIBs).
- Will publish updated MIB IDs after this meeting.
- Need to extend conformance, performance tables, consider
how to expose more information about hops (tunnel heads,
tails and intermediates), etc.
- Need to determine whether or not discriminated unions
should be supported.
- Multiple objects from multiple label types
- Also need to know who has done an implementation of these
- Bert Wijnen: reaffirmed that need people to review mibs
even if they are not MIB doctors to ensure it represents
model of technology as needed.
- Kireeti Kompella: who reads the MIBs? - How do you make
people read MIBs? This is part of progress of the protocol
the documents. We need feedback to know if we are going in
the right directions.
GMPLS-based Recovery (Dimitri Papadimitriou)
Dimitri Papadimitriou provided a status on these i-d's (the
terminology, analysis and functional specification are
closed, the signaling needs to pass through a thorough
review after becoming a working group i-d), also pointed out
that these documents should be submitted to the IESG for the
Dec'03 milestone (as per charter).
- Adrian Farrel: count of approx. 8 read the drafts, and no-
one thought that the four drafts content overlapped e.g.
that four were too many - authors will ask for consensus
via mailing list.
ASON Signaling Requirements (Jerry Ash)
Jerry Ash presented an update status on the document (note:
the ver. under discussion is v04.txt) authors believe that
the document is ready for WG last call.
- Adrian Farrel: asked for count of who read document double
digits (more than 12), and support for going to WG last
call also shows double digits
- Adrian Farrel: ask if there are any objections.
- Jonathan Sadler: noted that the proposed definition of
call segment is still being progressed at the ITU (thus
should add this clarification in the document).
ASON Interworking (Lyndon Ong)
- Dimitri Papadimitriou: pointing to the penultimate slide,
commented you have identified issues, and on mailing list
were identified issues, and the authors have not responded
to these issues. Proposing the RFC 3474 as an "existing
RFC", but what is rationale for the CCAMP WG to deal w/ it
since it has an informational status at IETF.
- Lyndon Ong: it is there, and its tied with an ITU standard
- Dimitri Papadimitriou: technically speaking the first
issue we have to deal with is the usage of TNAs, but do
not see any real rationale for introducing them.
- Kireeti Kompella: issue I have is that we have an existing
document, RFC 3473, which is by definition the baseline
from which we are building on and then there is no
- Kireeti Kompella (Question for CCAMP): do we want to
interwork? Before we go into technical issues, we need to
answer question what do we do with RFC 3474? What is its
exact status? and then do we look at it for interworking?
- Lou Berger: what is confusing me and is where do we start
from: we have two RFCs - we have a long history of having
RFCs which are informational, but there is a difference
between a proposed standard and an informational RFC.
Here we're using ITU as rationale, this is why we are
having this discussion. As info RFC 3474 is an ITU work,
what we should be doing is coordinating and ask what does
it take to support it? Is the info RFC 3474 the only way
for achieving these requirements? Do we have to support
RFC 3473? The response is yes, since RFC 3473 has the
standing here in CCAMP.
He also wanted to know whether this work is representative
of a number of people in ASON, or is it the work of a few
people looking to do things a different way.
- Lyndon Ong: ITU 7713.2 and RFC 3474 are tied to each other
and they do not represent a minority view.
- Bert Wijnen: Substantiated this point.
- Lyndon Ong: Pointed out that this group could either start
with what has been defined already or start over. The
latter approach is more likely to produce divergence
rather than convergence.
- Deborah Brungard: this draft is missing consideration of
backward compatibility aspects with RFC 3473 - CCAMP WG
needs to consider RFC 3473 compatibility, not just
compatibility with info RFC 3474.
- Bert Wijnen: CCAMP WG needs to identify to ITU any fatal
flaws with info RFC 3474, and work with them.
- Kireeti Kompella: What will the IETF and ITU will if there
is some vendor with 5K implementations with the "fatal
He then observed that the discussion is taking far longer
than he had expected and asked that people at the mike
should be the last and further discussion clearly needs to
take place on the list.
- Malcolm Betts: the emphasis should be on need to move
forward and the definition of future capabilities.
- Lou Berger: CCAMP WG needs to look at info RFC 3474 and
identify the flaws, and work on a standard ASON GMPLS
solution here in CCAMP.
RSVP-TE ASON (Dimitri Papadimitriou)
- Lyndon Ong: disagrees with conclusion that the RFC 3474 is
not backward compatible with RFC 3473, e.g. RFC 3474 does
not require an intermediate node to support RFC 3474. And
that RFC 3474 does not address all the requirements is
immaterial, it was done at a certain time, and new
capabilities will need to be added.
- Dimitri Papadimitriou: you say need convergence but the
3474 extensions do not address the needed functionality
and the analysis provided in the appendix of this i-d
shows where the backward compatibility issues are if this
- Kireeti Kompella: take to list to discuss technical
arguments if RFC 3474 compliant or not, and if it meets
requirements or not. Will need to liaise with ITU
- Lyndon Ong: need to take into account that this method is
an already agreed ITU standard so question is why diverge
from the RFC 3474 solution.
- Kireeti Kompella: we also have to agreed on a standard
from the IETF side. Lets look at technical arguments on
3474, and see how to get to an end point.
MPLS Crankback (Adrian Farrel)
Adrian Farrel gave a short status update
- The draft needs to deal with the fact that there are not
enough flag bits in the Session Attributes object.
- The authors need to define logical grouping of TLVs,
remove the unnumbered component link IF_ID TLV from this
draft (because it is more generally applicable).
- Adrian Farrel:
ask who read the draft: good showing
ask to become a WG document: no dissent
- Kireeti Kompella: good support, should keep as separate
document for now although will probably be part of an ASON
"boxed set" We need feedback, Adrian will lead the
discussion via mailing list.
ASON Routing Requirements (Deborah Brungard)
- Liaison statement to ITU-T
- Zafir Ali: question if requirements are from service
providers (and not only vendors)
- Deborah Brungard: yes, providers and vendors participate
- Kireeti Kompella: we will start with the requirements from
ITU and prioritize them
- Deborah Brungard: important to understand terminology and
dialogue on understanding requirements with ITU.
- Kam Lam: will include G.7715.1 ?
- Deborah Brungard: yes
- Kireeti Kompella: waxed philosophical about the advantages
of striving to attain a state of boredom in the CCAMP WG.
- Adrian Farrel: Ron Bonica missed the submission deadline.
This is now on our charter so we need to pay attention.
Arthi Ayyangar - draft-ayyangar-inter-region-te-01.txt
- Discussed some issue with terminology - specifically
the over-loading of certain terms. She also talked
about possible strategies related to the duplication
of work in this and - the next - draft.
JP Vasseur - draft-vasseur-inter-as-te-01.txt
- He gave a brief status/history of the draft - what
charter item it attempts to address, how long they
have been working on it, etc.
- Dimitri Papadimitriou: Wanted to know whether or not we
would first consider whether or not LSR PCS should be
done before we consider how to do it.
- JP Vasseur: Suggested that this is a question, among
others, for the list.
- Arthi Ayyangar: Pointed out that the discussion needs to
focus on requirements before focus on a solution.
- Kireeti Kompella: Looking for a single document for
both models of signaling. He would also like this work
to include applicability for each different model.
- Arthi Ayyangar: Please clarify what set of drafts are
- Kireeti Kompella: Provided a breakdown of the documents
and the issues with where the work might be done on each
part of the set.
- Adrian Farrel: Can we have a date by which the combined
draft will be produced? He would like the groups involved
to take some time this week to get started.
- JP Vasseur (with good grace): January 16th 2004
- Kireeti Kompella (summarizing):
Should we have inter-area and inter-as as one draft and
include both solutions and show when applicable the
different solutions for different scenarios? -> yes
Should loose re-optimization go to CCAMP? It is related
work. -> need to discuss this among the chairs/ADs
Should this item address both packet and non-packet?
Concerning PCS signaling need to discuss among chairs/ADs,
and if agreement to add in the charter, need for re-
chartering, and then address the technical aspects
Communication of LSP Alarm Information (Lou Berger)
He said they do currently have some issues. A key issue is
that the standard alarm information defined by Telcordia
and ITU are mostly in the form of strings.
- Kam Lam: points to X.733/X.736 and M.3100, will discuss it
- Kireeti Kompella: Might need to update charter before
considering this item, chairs and ADs need to discuss.
Also, not too many read draft, need to start thread on
email list, need to get carriers to speak up.
GMPLS Signaling for L2 LSP services ( Dimitri Papadimitriou)
Dimitri talked about some work that they have recently
started for L2SC in GMPLS. This work does not include
use of PW over PSN.
- Ask who read the draft: 12 at least
- Ask who feel it should be an item CCAMP should work on:
12 at least
- Kireeti Kompella: Pointed out that this work is not in the
CCAMP charter and it may be difficult to add it because
there is no focus in the IETF for L2 services.
- Kireeti Kompella: not in CCAMP charter to do technology-
specific work, need to discuss if it can go in charter.
Already have SDH and G.709.
And we need to check is there no other layer 2 specific
work in other IETF groups, then probably could be in CCAMP
- Dimitri Papadimitriou: point out that even if there are no
layer 2 specific documents, RFC 3473 and GMPLS
Architecture covers L2SC LSP concept (in particular, ATM
and FR). So this should equally be covered by the existing
He also asked what he would need to do to strengthen the
argument for getting this work accepted by the ADs.
- Kireeti Kompella: Consensus is a powerful argument.
We need to discuss this with the ADs before moving forward
Component Link Recording and Resource Control for GMPLS
Link Bundles (Zafar Ali)
- ask who read the draft: 8 to 10
- ask who feel it should become a WG document: 8 to 10
- Kireeti Kompella: concerns about what it is trying to
fix. Need reason to put this in ERO, so need to understand
why we want to put this in the ERO
- Adrian Farrel: would like to hear from providers on need
to use this before trying to adopt it. Take it to the list
Requirements for time-bounded notification of faults
- Adrian Farrel: the discussion should be taken to the list
in order to agree on requirements before looking at
*** Meeting in adjourned ***