Document: draft-ietf-l3vpn-2547bis-mcast-bgp-07 Reviewer: Spencer Dawkins Review Date: 2009-09-17 (oops!) IETF LC End Date: 2009-09-08 IESG Telechat date: (not known) Summary: This draft is almost ready for publication as a Proposed Standard. I have some comments, mostly 2119 language plus some sentences that didn't parse and seemed important. Abstract This document describes the BGP encodings and procedures for exchanging the information elements required by Multicast in MPLS/BGP IP VPNs, as specified in [MVPN]. Spencer (nit): I'd expect the RFC Editor to remove a reference in the Abstract section. 2. Introduction This document describes the BGP encodings and procedures for exchanging the information elements required by Multicast in MPLS/BGP IP VPNs, as specified in [MVPN]. This document assumes a thorough familiarity with procedures, concepts and terms described in [MVPN]. This document defines a new NLRI, MCAST-VPN NLRI. The MCAST-VPN NLRI Spencer (nit): I'd expect the RFC Editor to ask that abbreviations be expanded on first use. is used for MVPN auto-discovery, advertising MVPN to I-PMSI tunnel binding, advertising (C-S, C-G) to S-PMSI tunnel binding, VPN customer multicast routing information exchange among PEs, choosing a single forwarder PE, and for procedures in support of co-locating a C-RP on a PE. 5. PMSI Tunnel attribute When a router that receives a BGP Update that contains the PMSI Tunnel attribute with its Partial bit set determines that the attribute is malformed, the router SHOULD treat this Update as though all the routes contained in this Update had been withdrawn. Spencer (minor): why SHOULD in this case, and not MUST? (Note that similar text occurs throughout this section) An implementation MUST provide debugging facilities to permit issues caused by malformed PMSI Tunnel attribute to be diagnosed. At a minimum, such facilities SHOULD include logging an error when such an attribute is detected. Spencer (minor): If debugging facilities are a MUST, shouldn't the minimum debugging facility be a MUST? :-) Or are there well-understood reasons ("MUST do X unless") that could be provided? (Note that similar text occurs elsewhere in the draft) The PMSI Tunnel attribute is used in conjunction with Intra-AS I-PMSI A-D routes, Inter-AS I-PMSI A-D routes, S-PMSI A-D routes, and Leaf A-D routes. 7. VRF Route Import Extended Community If a PE uses Route Target Constrain [RT-CONSTRAIN], the PE SHOULD advertise all such C-multicast Import RTs using Route Target Constrains (note that doing this requires just a single Route Target Constraint advertisement by the PE). This allows each C-multicast route to reach only the relevant PE. To constrain distribution of the Route Target Constrain routes to the AS of the advertising PE these routes SHOULD carry the NO_EXPORT Community ([RFC1997]). Spencer (minor): I'm not sure I understand why these are SHOULDs and not MUSTs. (Note that similar text occurs elsewhere in the draft) I note that you give reasons why the similar text in Section 8 is not a MUST, using this explanation: "Note that if non-segmented inter-AS P-tunnels are being used, then the Intra-AS I-PMSI routes need to be distributed to other ASes and MUST NOT carry the NO_EXPORT community." 8. PE Distinguisher Labels Attribute The Next Hop field of the MP_REACH_NLRI attribute of the route SHOULD be set to the same IP address as the one carried in the Originating Router's IP Address field. Spencer (minor): I'm not sure why this is a SHOULD and not a MUST, but if it is a SHOULD, I'd expect to see some guidance about what happens if it's not observed - does the mechanism still work? 10. Non-congruent Unicast and Multicast Connectivity It is possible to deploy MVPN such the multicast routing and the Spencer (minor): is this "such that"? This is probably a nit but I'm having to guess at the parsing. unicast routing are "non-congruent". For instance, the CEs may be distributing to the PEs a special set of unicast routes that are to be used exclusively for the purpose of upstream multicast hop selection, and not used for unicast routing at all. (For example, when BGP is the CE-PE unicast routing protocol, the CEs may be using SAFI 2 ("Network Layer Reachability Information used for multicast forwarding" [IANA-SAFI]), and either IPv4 or IPv6 AFI to distribute a special set of routes that are to be used for, and only for, upstream multicast hop selection.) In such a situation, we will speak of the MVPN as having two VRFs on a given PE, one containing the routes that are used for unicast, the other containing the unicast routes that are used for upstream multicast hop (UMH) selection. We will call the former the "unicast routing VRF", and the latter the "UMH VRF" (upstream multicast hop VRF). 11.4. C-multicast routes aggregation Further a BGP receiver, that receives multiple such routes with the same NLRI for the same C-multicast route, will potentially create forwarding state based on a single C-multicast route. As per the procedures described in Section "Receiving C-Multicast Routes by a PE", this forwarding state will be the same as the state that would have been created based an other route with same NLRI. Spencer (nit): I'm pretty sure this is "based on another route" - but it doesn't parse for me without help. 12.1. Originating S-PMSI A-D routes In all of the above cases an implementation MUST allow to modify the default via configuration. Spencer (minor): sorry, this doesn't parse. I'm guessing it's missing a noun, but I can't suggest text. 16.1.1. Dampening withdrawals of C-multicast routes + Until the withdrawls are actually sent, multicast traffic for the Spencer (nit): s/withdrawls/withdrawals/g, because I saw other occurances in the document... C-multicast routes in question will be continued to be transmitted to the PE, which will just have to discard it. Note that the PE may receive useless (multicast) traffic anyway, irrespective of dampening of withdrawals of C-multicast routes due to the use of I-PMSIs. 17. Security Considerations The mechanisms described in this document could re-use the existing BGP security mechanisms. Specer (minor): "could" seems awfully nebulous for security considerations :-) I'm guessing you could delete this sentence, but it needs help of some sort. 18. IANA Considerations Spencer (minor): I see from the ID tracker that IANA has already asked for clarification on registration procedures ... :-)