Notes from the MPLS WG meeting at the Minneapolis IETF ====================================================== Tuesday November 18 0900-1130 Salon D 1. Agenda bashing ----------------- The agenda was accepted with the changed that agenda item #3 was moved to the MPLS-TP session on Wednesday. 2. Working group status --------------------------------- http://www1.tools.ietf.org/wg/mpls/ In Dublin the MPLS working group were asked to start a work to recharter - this work has not resulted in any firm results yet. We have three new RFCs since Dublin: - RFC 5530 "A Link-Type sub-TLV to Convey the Number of Traffic Engineering Label Switched Paths Signalled with Zero Reserved Bandwidth across a Link" - RFC 5331 "MPLS Upstream Label Assignment and Context-Specific Label Space" - RFC 5332 "MPLS Multicast Encapsulations" We have no documents in the RFC editors Queue. The following document are processed by the IESG - mpls-ldp-igp-sync - mpls-te-scaling-analysis - mpls-p2mp-te-mib - mpls-ldp-capabilities The following documents are ready for wg last call - draft-ietf-mpls-cosfield-def-07.txt - draft-ietf-mpls-rsvp-upstream-03.txt - draft-ietf-mpls-ldp-upstream-03.txt We have adopted one new working group document recently - draft-mpls-mpls-ip-options-01 We will have a new Boiler Plate for Internet Drafts that will be mandatory staring December 16th 2008. We have some MPLS related work in other groups - Routing Area –router alert considerations - BMGW – mpls forwarding method MPLS-TP stuff - meeting on Wednesday, joint with CCAMP and - there will be no MPLS-TP interim in Malta, but we are looking into alternatives 3. Multiprotocol Label Switching (MPLS) Traffic Engineering Management Information Base for Fast Reroute ------------------------------------------------------------ Lou and JP Lou presented LSP Reroute draft. The changes to this draft is minor as compared with the situation Dublin. Draft Alignment – authors believe there is good alignment between the three drafts for MPLS, but CCAMP will need to address an open issue. Next steps – an updated draft will be issued and then it should ready for publication. JP presented 3209 PathErr draft and soft preemption drafts. Agrees with Lou that there is a good alignment between the three drafts. Loa will package them as a group and send them to the IESG. Need to have updated drafts and let Loa know when ready to send. 5. Proxy LSP Ping ------------------ George Swallow presented the draft. Comments has been solicited on the working group mailing list, but no were comments receeved. The author believe that the draft is ready for working group last call and that an early security would be appropriate. The document will working group last called. 6. Mechanism for performing LSP-Ping over MPLS tunnels ------------------------------------------------------- Ice presented the changes to the draft since Dublin. There are simulations planned based on this draft, but the work has just started, but the working group will be kept updated. 8. Framework and Requirements for Composite Transport Group (CTG) ------------------------------------------------------------------ Ning So The draft was presented by Ning So and since this draft was presented in several working groups during the meeting in Minneapolis he was instructed to give to specifically address why this draft should be picked up by the MPLS working group. ome issues has been identified in MPLS deployments with the existing solutions on utilization of parallel links. The mechanism specified in this draft addresses this problem. This mechanism is intended to work with packet based MPLS as well as transport mechanisms applicable to CCAMP, (e.g., lambdas, optical links, etc.) Ning So promised to put further work into the requirements section. The chairs pointed to the possibility that this work could be taken up by the routing area (rtgwg). The working group chairs take the action point to bring this to the attention of ADs and wg chairs. 9. Ethernet MAC Destination Address for Multicast MPLS ------------------------------------------------------ Thomas Thomas presented the draft. The aim this draft is to the clarify when an LSR should use Ethernet multicast and uni-cast addresses, and when a MPLS encapsulated packet is considered a uni-cast or multicast. This is considered to be an update of RFC5332. The discussion focused on whether the draft addresses a real problem. Since the discussion did not converge the said that the discussion should continue on the list. 9. Mechanism for performing LSP-Ping over RSVP protection paths ---------------------------------------------------------------- Ice Ice presented the document that proposes a simple method to map IP multicast tree to MPLS p2mp or mp2mp LSPs. The authors requests comments on the draft.expect comments 11. Using mLDP through a Backbone where there is no Route to the Root --------------------------------------------------------------------- Ice Ice presented the draft that specifies procedures which enable a point tp miltipointMP LSP to be constructed through a BGP-free core. In these procedures, the root node address is temporarily replaced by an address which is known to the intermediate nodes. Comments from the working group are requested. 12. Signaled PID When Multiplexing Multiple Payloads over RSVP-TE LSPs --------------------------------------------------------- Zafar The document identifies that there are deployment scenarios where an RSVP-TE LSP carries multiple types of payloads. This could lead to ambiguity of which PID to put in the label request object. The document offers a method to clarify this ambiguity. The discussion focused on whether this is a real problem or not. 12. Signaling RSVP-TE P2MP LSPs in an Inter-domain Environment -------------------------------------------------------------- Zafar The draft says that current specifaction leaves certain things open when it comes to establishing p2mp LSP in a multidomain envronment. Discussion foucused on whether the existing solutions are enough or needs to extended. The discussion should continue on the MPLS wg mailing list. 13. The Use of Entropy Labels in MPLS Forwarding ------------------------------------------------- Kireeti The draft describes and define the concept of "entropy labels" and how they can be used to improve load balancing. Entropy labels is not used for forwarding, they are not signaled and the sole pupose improve load balancing. Discussion to continue on the mailing list. 14. End of Tuesday meeting ------------------------- Minutes for the joint working group meeting on MPLS-TP ====================================================== Wednesday, November 19, 2008, 13.00 - 15.00 Salon AB 1. Agenda bashing ----------------- This meeting is joint meeting with mpls, ccamp,pwe3 and l2vpn working groups. Minutes wioll be published as part of the MPLS working group minutes. Note: The MPLS / GMPLS security framework that is now ready for WG last call. Please read carefully and comment. The agenda was approved as proposed. 2. Status of the MPLS-TP work ------------------------------ Loa presented the status of the MPLS-TP work. The mailing list to use for discussion on mpls-tp is: mpls-tp@ietf.org (normal subscribition procedures) Find current MPLS-TP documents through the ID-tracker: https://datatracker.ietf.org/idtracker/ put "mpls-tp" in the filename field. The document on renaming the EXP field in an MPLS label stack entry has been wrapped up and is now in IESG review. The MPLS-TP is progressing a good speed. 4 MPLS-TP documents are progressed to WG documents and another set of four will soon be polled to become working group documents. We will not have an interim meeting in Malta meeting, bur are looking for an opportunity to meet with editors and authors of documents that needs to be progressed. We will send a liaison to the ITU-T with information om the progress, this liaison will request a definitive list of documents that needs to be RFCs in October 2009 and ask for review on the four working group documents. 3. An MPLS-TP Rosetta Stone ---------------------------- Huub Huub presented the work on a “Rosetta Stone” draft. There is a first version is posted and then everyone can comment on the mpls-tp list. 4. MPLS-TP Requirements ------------------------ Ben Ben presented the changes from October MEAD meeting and the outstanding work. He also stressed that now is the time to identify requirements that are missing and help reword introduction. There is consensus to accept this document to working group document. 5. A Framework for MPLS in Transport Networks ---------------------------------------------- Matthew Matthew presented the work that has been done since Dublin. There is still some issues that we need to work out on the list. In particular it was pointede out that we have overlap issues between several document, e.g. between requirement and framework documents. We need to find a balance between referencing another document and to allow for enough context to make a particular document readable. In particular it is important that the MPLS-TP framework documents gives enough inforamtion to be possible to read stand alone and used as a general overview of MPLS-TP. There is consensus to accept this document to working group document. 6. Requirements for OAM in MPLS Transport Networks --------------------------------------------------- Martin Martin presented and the progress since Dublin. The main issue discussed in this meeting was the different compatibility issues. There are two potential sources for OAM functionality, what exists within the current MPLS/GMPLS standards, and what operators today frequently use. We should make sure that existing tools are used whenever possible, extended and improved if necessary. There is consensus to accept this document to working group document. 7. MPLS-TP OAM Framework and Overview -------------------------------------- Italo Italo presented the OAM framework drafts and updates since Dublin. There are still open issues, we need to address MPLS-TP point to multipoint and label stacking, and we need to define a terminology that captures protection and OAM aspects, e.g. to form a good basis for documents as the Survivability draft. We need to set an agressive target for this document and be reayd to move it to working group document by the end of January. We need OAM deployment scenarios for MPLS-TP OAM to understand the scaling capabilities. The discussion will continue on the mpls-tp mailing list. 8. MPLS Generic Associated Channel ----------------------------------- Matthew Matthew presented the updates and changes since Dublin. There is still some work on capabilities and terminology. Discussion focused on what the capabilites aof the Generic Associated Channel and what mechanisms should be supported. The channel is supported for both LSPs and PWs. There is consensus to move the document to WG draft and we intend to go to working group last call soon. 9. MPLS TP Network Management Requirements -------------------------------------------- Eric Eric presented the draft and the chneges since DUblin. It is a goal to progress the draft to working group document before the IETF 74. This requirement document will be followed bu a framework document that describes managed objects without saying how to manage them or what protocol to use in terminology that applies to both ITU and IETF. 10. MPLS-TP Control Plane Framework ------------------------------------ Lou Lou reported on the Control Plane work, no draft has been published yet. There is an document outline that has been circulated between editors and co-authors and which gives an idea what will go into it. The control plane overlaps with some of the other MPLS-TP documents, e.g. the protection and survivability documetns. We need to coordinate between these documents. 11. Report on analysis of the MPLS-TP ring protection requirements ------------------------------------------------------------------ John John reported from an actvity to nail down the the MPLS-TP requirements for ring protection. The tentative outcome of this study is that the delta between the generic protection requirements and ring protection are extremely small. This does not motivate producing specific ring protection requirements and framework documents. The discussion indicate that there is still some way to go before we have concluded the study. To do so we need get the concrete 11. MPLS-TP OAM Analysis ------------------------- Nurit This agenda item was not addressed due to lack of time. 12. Multiprotocol Label Switching Transport Profile Survivability Framework ---------------------------------------------------- Nurit This agenda item was not addressed due to lack of time. 13. Operating MPLS Transport Profile LSP in Loopback Mode ---------------------------------------------------------- http://www.tla-group.com/~loa/draft-boutros-mpls-tp-loopback-01.txt Sami This agenda item was not addressed due to lack of time. 13. End of meeting ------------------ See you all in San Francisco next year!