2.5.6 Layer 1 Virtual Private Networks (l1vpn)

NOTE: This charter is a snapshot of the 63rd IETF Meeting in Paris, France. It may now be out-of-date.

Last Modified: 2005-06-29


Hamid Ould-Brahim <hbrahim@nortel.ca>
Adrian Farrel <adrian@olddog.co.uk>
Tomonori Takeda <takeda.tomonori@lab.ntt.co.jp>

Routing Area Director(s):

Bill Fenner <fenner@research.att.com>
Alex Zinin <zinin@psg.com>

Routing Area Advisor:

Alex Zinin <zinin@psg.com>

Mailing Lists:

General Discussion: l1vpn@ietf.org
To Subscribe: https://www1.ietf.org/mailman/listinfo/l1vpn
Archive: http://www.ietf.org/mail-archive/web/l1vpn/index.html

Description of Working Group:

No description available

Goals and Milestones:

Sep 05  Submit first Internet Draft of L1VPN framework
Sep 05  Submit first Internet Drafts of basic mode specifications
Dec 05  Submit first Internet Drafts of MIB modules for basic mode
Apr 06  Submit basic mode specifications to IESG for publication as Proposed Standard
Jun 06  Submit first Internet Drafts of enhanced mode specifications
Aug 06  Submit MIB modules for basic mode to IESG for publication as Proposed Standard
Dec 06  Submit enhanced mode specifications to IESG for publication as Proposed Standard
Dec 06  Submit L1VPN framework to IESG for publication as Informational RFC
Aug 07  Submit MIB modules for enhanced mode to IESG for publication as Proposed Standard
Dec 07  Recharter or disband

No Current Internet-Drafts

No Request For Comments

Current Meeting Report

Layer 1 Virtual Private Networks (l1vpn)

TUESDAY, August 2 at 1630-1800

CHAIRS: Hamid Ould-Brahim <hbrahim@nortel.com>
Adrian Farrel <adrian@olddog.co.uk>
Tomonori Takeda <takeda.tomonori@lab.ntt.co.jp>

1) Group Admin
- Blue sheets, minute takers, admin, agenda bashing

Hamid Ould-Brahim went over the agenda.

2) Welcome to L1VPN
- Welcome and context

- Alex Zinin: You asked for it and you got it. Looking for rapid progress on protocol aspects from the group. Three chairs for L1VPN in tradition of VPN working groups (joke).

3) Goals, objectives and milestones of the WG chairs
- Scope, milestones, priorities

Tomonori Takeda went over the charter. Stated that the basic mode is the initial focus.

- Adrian Farrel: Reminders to WG on Goals and non goals. You can help reaching Milestones and you will be Adrian's friend.

4) SG13 status and Recommendations
Marco Carugi

Marco Carugi went over the slide, summarized the L1VPN working in the ITU and the interaction with the IETF to date.

5) IETF framework draft
Tomonori Takeda

Tomonori Takeda went over the slide, summarizing mode/models and requirements.

- Hamid Ould-Brahim: Who has read ? some show of hands
- Adrain Farrel: Please read this I-D as this is basic to work for going forward.

6) Generalized VPN draft
Hamid Ould-Brahim

Hamid Ould-Brahim went over the slide, summarizing the models and functionalities.

7 )GMPLS Overlay draft Adrian Farrel (5 mins)

In RFC Editor Queue currently. Adrian noted that the draft addresses both gmpls core networks and non-gmpls, here we are addressing gmpls core networks. L1VPN model from overlay draft is used as model in applicability draft.

8) Applicability and Solutions
Dimitri Papadimitriou

Dimitri Papadimitriou went over the slide, explaining the purpose of the draft and basic mode solutions.

- Adrian Farrel: Are there missing requirements?
- Dimitri Papadimitriou: Who read the draft? - a few show of hands
- Hamid Ould-Brahim: Suggest read the draft and provide comments.

- Lou Berger: One Comment. How do you identify the VPN specific addresses across the provider network, address translation, address resolution?
- Dimitri Papadimitriou: The procedures will be part of detailing the operations.
- Lou Berger: That implies a routing solution?
- Hamid Ould-Brahim: Not necessarily. The membership identity depends on the model. Can be implicit or explicit.
- Lou Berer: What about the Ports? Addresses may be different in different solutions.
- Yakov Rekhter: We have auto discovery. As part of auto discovery we can do it.

9) Solutions discussion (Basic Mode only)

Hamid Ould-Brahim went over the slide, explaining dates and targets.

- Bijan Jabbari: If you had a Pseudo wire and you are running SDH/SONET on it, does this working group preclude this scenario?
- Hamid Ould-Brahim: We are emphasizing l1vpn and gmpls.
- Bijan Jabbari: If you use GMPLS for it? Is it an L1VPN?
- Adrian Farrel: If it is L1VPN service over GMPLS-based network, would be OK.
- Alex Zinin: Is there a document on that?
- Bijan Jabbari: No.
- Alex Zinin: Need to understand what's different. Need a draft to understand what you are looking for.
- Kireeti Kompella: We discussed overlap of PWE3 and L2VPN before. We separated it. PWE3 is set up PW. L2VPN is beyond that. So they are different. Here it is a cleaner separation as we are not impacting how it is transported over the network.
- Igor Bryskin: Agree with Kireeti. L1VPN is far wider than PW, and it's orthogonal. We are talking about the control plane connecting circuits. It is pretty much orthogonal.

- Kireeti Kompella: Speaking of running before we can walk. Things like scheduling. First I discover. Then I make the connection. I may not want to make the connection right away. We have not done anything that is time bounded. That is a reasonable question. Adding a time dimension.
- Bijan Jabbari: We do have a dedicated resource for a schedule so many days ahead of time. I do not know whether the scheduling problem fits. I'm not sure whether the scheduling problem fits with the charter of the WG. If you could mention in the framework draft.
- Adrian Farrel: Please write it.
- Otto Kreiter: Research network in Europe there is a big demand. Perhaps we can introduce the scheduling.
- Adrian Farrel: Can you write some text proposal on this requirement?
- Igor Bryskin: Can be answered simply, I am buying not a connection, I am buying a connectivity, so should be a guarantee, that the network will be available.
- Kireeti Kompella: The problem of knowing if there is enough capacity at some scheduled time in the future we have not looked at, need to add the time dimension.

- Mark Townsley: Adding time dimension is interesting. If this is proposed here to do time dimension, should also circulate to other groups, L2VPN, etc.
- Adrian Farrel: Do you think this type of requirement is greater relative to the size of the connection?
- Kireeti Kompella: Not so much greater with the size of the connectivity. It is the nature of the problem. L2VPNs I setup right away. In L1VPNs I set this up in the future.
- Bijan Jabbari: Bigger chance of aggregation in L2 and L3. The demand is not that big.
- Loa Andersson: There are such services in the MPLS world, the most stringent requirement is to get it to the customer on-time with tenths of secs, and to be able to delete again.
- Igor Bryskin: Connection vs Connectivity. Each will get the connection but what he needs is connectivity on demand.
- George Swallow: Much more complexity for managing to do scheduling, so may only be for large pipes that need this.
- Bijan Jabbari(?): There is info available on this.

- Himanshu Shah: MPLS forum has the MPLS UNI. This seems like that.
- Yakov Rekhter: IETF has it's own UNI, other SDOs can do UNI but it's their problem, first we should do IETF UNI.


Goals, objectives and milestones of the WG (chairs)
SG13 status and Recommendations
IETF framework draft
Generalized VPN draft
GMPLS Overlay draft
Applicability and Solutions
Target dates and milestones