Hello After reading the LISP+ALT (draft-ietf-lisp-alt-01) I have following comments, suggestions and questions. General comment: LISP+ALT draft lacks consistency with draft-ietf-lisp-ms-04, instead the term "configuration" is used quite broadly covering the area of the mentioned draft and leaving too much for a reader to fill in the caps. See below what I mean: Section 4 Sentence from alt-01: "Note that ETRs are not required to participate (or prevented from participating) in LISP+ALT; they may choose communicate their mappings to their serving LISP+ALT router(s) at subscription time via configuration." This is very true, but ETRs may also choose to use Map-Register messages that contain a list of EID-prefixes plus a set of RLOC as defined in draft-ietf-lisp-ms-04.txt section 4. The issue here is that this is not only subscription time event but a continuous (periodic to be more specific) process as defined in draft-lisp-ms. Section 5 Sentence "A LISP+ALT router near the edge learns EID prefixes originated by authoritative ETRs, either by eBGP peering with them or by configuration." This should be aligned with ietf-lisp-ms-04.txt to cover Map-registering. Section 5.2 "It also implies that an ETR that originates a prefix must maintain BGP sessions with all ALT routers that are configured to originate an aggregate which covers that prefix. " This I find contradictory to the target to keep ETR a simple CPE device as it insists on the use of BGP peering. Section 5.4 "It receives Data Probes and Map-Requests only over GRE tunnel(s) to its "upstream" LISP+ALT router(s) and responds with Map- Replies for the EID prefixes that it "owns". " Here the text is about ETR. I guess that is should say "receives Data Probes and Map-Requests ... FROM its upstream... " Section 6.2 Sentence "these LISP+ALT router(s) the ETR must route Map-Requests and Data Probes to the ETR and contain configuration (in effect, static routes) for the ETR's ..." The first occurrence of "the ETR" looks like a typo. If not more clarification in the text would improve its readability. Notice a again the use of the term of "configuration" in this context. Not really sure if this is the similar configuration that was implied in the earlier part of the text. Section 8.1 "Not only does this greatly improve the scalability of the global routing system but it also allows improved traffic engineering techniques by allowing richer and more fine-grained policies to be applied." But how are the operator policies and the mapping system policies correlated? The network can only apply its existing BGP policies for data packets entering into its network, but not during the map resolution phase? Section 9.1 "As in the case above, the ETR is connected to LISP+ALT router(s) using GRE tunnel(s) but rather than BGP being used, the LISP+ALT router(s) are configured with what are in effect "static routes" for the EID prefixes "owned" by the ETR. " Maybe ETR chooses to use the Map-Register message as in draft-eitf-lisp-ms, or is this what is meant by static routes? In case the ETR happens to have connections to multiple LISP+ALT routers in different parts of the ALT hierarchy how does the ETR know where to register/configure what EID prefix? Best regards Hannu
Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.