MPLS Working Group Loa Andersson Internet-Draft Informational George Swallow Expiration Date: April 2003 Cisco Systems 26 November, 2002 The MPLS Working Group decision on MPLS signaling protocols Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026 [RFC2026]. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. For potential updates to the above required-text see: http://www.ietf.org/ietf/1id-guidelines.txt Abstract This document documents the consensus reached by the MPLS Working Group within the IETF to focus its efforts on "RSVP-TE: Extensions to RSVP for LSP Tunnels" (RFC3209) as the MPLS signalling protocol for traffic engineering applications and to undertake no new efforts relating to "Constraint-Based LSP Setup using LDP" (RFC3212). -When this document has been approved by the IESG, the following note will be added to the abstract: "The recommendations of chapter 6 have been accepted by the IESG" Conventions used in this document INTERNET-DRAFT draft-andersson-mpls-sig-decision-02.txt 08.11.02 Andersson Expires April 2003 [Page 2] The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC-2119 [RFC2119]. Contents 1. Introduction ...................................................... 2 1.1 Objectives of document ........................................ 2 1.2 Nomenclature .................................................. 3 2. Background ........................................................ 3 3. CCAMP implementation study ........................................ 4 4. MPLS Working Group discussion ..................................... 4 4.1 Phase 1 ....................................................... 4 4.2 IETF process .................................................. 5 4.3 Relationship to other standards organizations ................. 5 4.4 Phase 2 ....................................................... 6 5. MPLS Working Group consensus ...................................... 7 6. Recommendation to the IESG ........................................ 8 7. Security considerations ........................................... 8 8. IANA considerations ............................................... 8 9. References ........................................................ 9 9.1 Normative ..................................................... 9 9.2 Non-normative ................................................. 9 1. Introduction 1.1 Objectives of document This document documents the MPLS Working group consensus to continue to develop RFC3209 [RFC3209] as the signalling protocol for MPLS signaling for Traffic Engineering applications. This document also documents the MPLS working group consensus to not undertake any new work related to RFC3212 [RFC3212], e.g. there are no INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 3] plans to progress RFC 3212 beyond proposed standard. No other actions are taken relative the document status of RFC3212 [RFC3212] or RFCs that specify extensions to RFC3212. In section 6 we put forward the decision we believe the IESG should take based on the consensus in MPLS working group on this issue. All other sections is a documentation of the consensus process. 1.2 Nomenclature This document uses the term "CR-LDP related working group drafts" to refer to a group of Internet Drafts that specify changes or extensions to [RFC3212] and the term "CR-LDP related RFCs" to discuss the group of RFCs that specify the protocol and the applicability of [RFC3212]. The CR-LDP related working group drafts are: "Multi Protocol Label Switching Label Distribution Protocol Query Message Description" [QUERY] "Improving Topology Data Base Accuracy with Label Switched Path Feedback in Constraint Based Label Distribution Protocol [FEED] "Signalling Unnumbered Links in CR-LDP" [UNNUM] "Fault Tolerance for the Label Distribution Protocol (LDP)" [FT] "Generalized MPLS Signaling - CR-LDP Extensions" [GEN] "Generalized Multi-Protocol Label Switching Extensions for SONET and SDH Control" [SONET] "Generalized MPLS Signalling Extensions for G.709 Optical Transport Networks Control" [G709] "Generalized Multiprotocol Label Switching Extensions to Control Non-Standard SONET and SDH Features" [SDH] CR-LDP related RFCs The CR-LDP related RFCs are: RFC3212, "Constraint-Based LSP Setup using LDP" RFC3213, "Applicability Statement for CR-LDP" RFC3214, "LSP Modification Using CR-LDP" No further updates of the CR-LDP related RFCs, beyond their current statuses are planned within the MPLS Working Group. 2. Background Very early (1997) in the MPLS standardization it was clear that a protocol would be needed that would enable providers to setup LSPs that took other information (e.g. various QoS parameters) into account. INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 4] Development of this type of signalling protocol took two different tracks: - extensions to RSVP for setting up MPLS tunnels [RFC3209] - extensions to LDP for setting constraint based LSPs [RFC3212] The motivation for the choice of protocol in both cases was straightforward. Extending RSVP-TE to do in an MPLS environment what it already was doing (handling QoS information and reserving resources) in an IP environment is comprehensible; you only have to add the label distribution capability. Extending a native MPLS protocol like LDP, which was designed to do label distribution, to handle some extra TLVs with QoS information is also not revolutionary. The MPLS group never reached a consensus on which way to go. Both protocols were progressed to proposed standard. 3. CCAMP implementation study An implementation survey of GMPLS implementations was published in June 2002 [GMPLS]. The survey includes responses from 22 different implementers. Twenty-one of 22 implementations include the GMPLS signalling based on [RFC3209], while only 3 include signalling based on [RFC3212]. 4. MPLS Working Group discussion 4.1 Phase 1 The GMPLS implementation report prompted questions asking if it was reasonable to have two different protocols for the same thing. The discussion was brought to the MPLS Working Group at the meeting in Yokohama in July 2002. After discussion at the meeting it was decided to "bring this to the list" and also invite comments from the other Sub-IP Area Working Groups. The following question sent to the mailing lists: "As there are issues with having two similar standards (potentially diverging), and it generates duplicate work in several IETF working groups, the question was asked whether we should make CR-LDP informational (which still make it available and possible to work with) and progress only RSVP-TE on the standards track." INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 5] The response to this question was largely positive, but some problems were immediately pointed out: - there are non-IETF standards which reference RFC3212. Taking CR-LDP off the standards track would cause un-necessary problems for those organisations and should be done only after co-ordinating with those organizations - there is, e.g. in RFC2026 [RFC2026], no documented process according to which a document on the standards track may be move to a status that is non-standards track Each of these arguments is by themselves strong and would have led to some reformulation of the proposal to move CR-LDP to informational. Moreover, in combination it was clear that the original proposal was not viable. On the other hand the support for doing additional development of CR-LDP as an IETF standards track alternative to RSVP-TE was extremely small. 4.2 IETF process The current IETF process for managing changes in RFC status does not include any information on how to move an existing standard track RFC to a non-standard track status, nor does it include a prohibition of such an action. It has been shown that such actions have been previously taken e.g. RFCs 2673 and 2874 were moved from Proposed Standard to Experimental. Though the cases are not exactly parallel to the MPLS signalling case it shows that the IETF and IESG are prepared to take such decisions given that the arguments are sufficiently strong. 4.3 Relationship to other standards organizations The relationship with other standard organizations is an important part IETF work. We are dependent on their work and they make use of our technology; each organization has their own area of expertise. It is therefore necessary that both sides handle their standards documentation in such a way that no unnecessary updates or revisions are introduced simply by sloppy handling of documents. Consequently we need to keep CR-LDP referenceable, i.e. on the standards track, for the foreseeable future. The implication of this is not that we need to progress it further, or need to undertake further work in the area. One implication however is that standards organizations which reference the document, need to be notified of our decision so that they (at their own pace) can change their references to more appropriate documents. It is also expected that they will notify us when they no longer have a need to normative reference to CR-LDP. INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 6] 4.4 Phase 2 Based on the feed back from this first discussion the question to the working group were reformulated as: "Should the MPLS WG focus its efforts on a signalling protocol for traffic engineering applications on RSVP-TE, and hence the WG effort with CR-LDP be discontinued? This would not involve any change in document status for CR-LDP, nor would it hinder continued individual contributions in the CR-LDP space. It would involve a change in the MPLS WG charter to reflect this." It was pointed out that "nor would it hinder continued individual contributions" is too weak. We actually discourage, while it is not prohibited, continued work in the CR-LDP area. That is the whole point with taking this decision. It was also pointed out that while it is quite acceptable to not accept further working group documents, it would also be appropriate to take the existing CR-LDP related working group Internet Drafts through the process to proposed standard or informational as intended. This is applicable to the following documents, since much of the work has already been completed on them: - in MPLS WG -- Multi Protocol Label Switching Label Distribution Protocol Query Message Description -- Improving Topology Data Base Accuracy with Label Switched Path -- Feedback in Constraint Based Label Distribution Protocol -- Signalling Unnumbered Links in CR-LDP -- Fault Tolerance for the Label Distribution Protocol (LDP) - in CCAMP WG -- Generalized MPLS Signaling - CR-LDP Extensions -- Generalized Multi-Protocol Label Switching Extensions for SONET and SDH Control -- Generalized MPLS Signalling Extensions for G.709 Optical Transport Networks Control -- Generalized Multiprotocol Label Switching Extensions to Control Non-Standard SONET and SDH Features Some of the documents listed above are not in themselves extensions to CR-LDP, but in one way or another are deemed to be "equally applicable to CR-LDP". For those documents it will be fully appropriate to progress them beyond proposed standard in the future if they meet the requirements. INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 7] RFCs that are extensions to CR-LDP, e.g., RFCs 3213 and 3214, will remain proposed standard documents. After this compromise was proposed a good consensus quickly formed supporting the proposal. Close to 90% of the people participating discussion said that they support or at least accept this outcome of the working group discussion. 5. MPLS Working Group consensus In a message to the working group (date) the working groups chairs stated that consensus had been reached on: - that the MPLS WG needs to focus its efforts on RSVP-TE (RFC 3209) as protocol for traffic engineering signalling. - that the Working Group will undertake no new work r5elated to CR- LDP. - that the WG charter should be updated to reflect this. - that the WG will recommend that CR-LDP (RFC 3212) remain a proposed standard. - that the WG will recommend that RFCs 3213 and 3214, which are closely related to CR-LDP, remain proposed standard. - that existing Working Group drafts related to or updating/changing CR-LDP will be progressed through the standards process to proposed standard or informational RFCs as appropriate. - that "the existing cr-ldp working group documents" are: -- Multi Protocol Label Switching Label Distribution Protocol Query Message Description -- Improving Topology Data Base Accuracy with Label Switched Path Feedback in Constraint Based Label Distribution Protocol Signalling Unnumbered Links in CR-LDP -- Fault Tolerance for the Label Distribution Protocol (LDP) -- Generalized MPLS Signaling - CR-LDP Extensions -- Generalized Multi-Protocol Label Switching Extensions for SONET and SDH Control -- Generalized MPLS Signalling Extensions for G.709 Optical Transport Networks Control -- Generalized Multiprotocol Label Switching Extensions to Control Non-Standard SONET and SDH Features INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 8] - that the MPLS working group will take on no new Working Group documents related to CR-LDP. - that the MPLS working group will entertain no efforts to promote CR-LDP beyond proposed standard. - that individual contributions related to CR-LDP area are not prohibited, but discouraged. - that a message will be sent to the relevant standards organizations notifying them of this change of focus on MPLS signalling protocols. 6. Recommendation to the IESG Based on the consensus in the MPLS working group we recommend the IESG to: - confirm the MPLS Working Group consensus to undertake no new work on CR-LDP and focus on RSVP-TE as signalling protocol for traffic engineering applications for MPLS, as described in this document - adopt as an IETF policy to refrain from entertaining work that intends to progress RFC-3212 or related RFCs beyond proposed standard - adopt as an IETF policy to refrain from entertaining new working group documents that are extensions to RFC3212 - review the IETF process with respect to management of documents that needs to be moved from standards track to any other status - publish this document as Informational RFC 7. Security considerations This document only discusses a refocusing of the MPLS Working Group work and consequently brings no new security considerations. 8. IANA considerations This document brings no IANA considerations. INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 9] 9. References 9.1 Normative [RFC2026] Bradner, S. "The Internet Standards Process -- Revision 3", RFC 2026, October 1996. [RFC2119] Bradner, S. "Key words for use in RFCs to Indicate Requirement Levels", RFC 2119, March 1997. [RFC3212] Jamoussi, B. et.al., "Constraint-Based LSP Setup using LDP" RFC3212, January 2002. [RFC3209] Awduche, D. et.al., "RSVP-TE: Extensions to RSVP for LSP Tunnels", RFC3209, December 2001. 9.2 Non-normative [RFC3213] Jamoussi, B. et.al., "Applicability Statement for CR-LDP", RFC3213, Jan 2002 [RFC3214] Jamoussi, B. et.al., "LSP Modification Using CR-LDP" RFC3213, Jan 2002 [GMPLS] Rekhther,Y and Berger,L, "Generalized MPLS Signaling - Implementation Survey" http://www.ietf.org/IESG/Implementations/MPLS-SIGNALING- Implementation.txt , June 2002. [QUERY] Ashwood-Smith P. and A. Paraschiv, "Multi Protocol Label Switching Label Distribution Protocol Query MessageDescription" Work in Progress, May 2002. [FEED] Jamoussi, B. et.al., "Improving Topology Data Base Accuracy with LSP Feedback in CR-LDP", Work in progress, May 2002. [UNNUM] Rekhter, Y., et.al., "Signalling Unnumbered Links in CR-LDP", Work in Progress, July 2002. INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02 Andersson Expires April 2003 [Page 10] [FT] Farrel, A., et.al., "Fault Tolerance for the Label Distribution Protocol (LDP)", Work in Progress, Sep 2002. [GEN] Ashwood-Smith, P. and Berger, L. (eds) "Generalized MPLS Signaling - CR-LDP Extensions", Work in Progress, Aug 2002. [SONET] Mannie, E and Papadimitriou, D., "Generalized Multiprotocol Label Switching Extensions forSONET and SDH Control" Work in Progress, Aug 2002. [G709] Papadimitriou, D. (ed), "Generalized MPLS Signalling Extensionsfor G.709 Optical Transport Networks Control" Work in Progress, June 2002. [SDH] "Generalized Multiprotocol Label Switching Extensions to Control Non- Standard SONET and SDH Features" Work in Progress, June 2002 Authors contacts: Loa Andersson email: loa@pi.se George Swallow Cisco Systems, Inc. 250 Apollo Drive Chelmsford, MA 01824 Voice: +1 978 244 8143 email: swallow@cisco.com INTERNET-DRAFT draft-andersson-mpls-sig-decision-03.txt 26.11.02