The MPLS WG Archive[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index][Thread Index][Author Index][Subject Index] AW: Encapsulating MPLS in IP or GRE
I like to refer to my presentation in Yokohama about building hierarchical LSPs (H-LSPs) by concatenating i.e. label-switching elementary LSP tunnels. I pointed out that elementary LSPs may be MPLS-LSPs as we know today or also Non-MPLS tunnels. Consequently the use of hierarchical LSP may a) eliminate the O(n**2)-problem, b)enable MPLS-VPNs spanning MPLS networks as well as non-MPLS networks. E.g. a single MPLS-tunnel between two PEs might require the employment of a p2p-H-LSP because it needs to be built by concatenating an mPLS-LSP and an IPSEC or GRE tunnel. IMO, we should discuss hierarchical LSPs at all. Maybe at first p2p-H-LSPs prior to mp2p resp. p2mp resp. mp2mp H-LSPs. Maybe at first H-LSPs which do label-switch MPLS-LSPs prior to those which do label-switch non-MPLS LSPs like IPSEC,GRE,IP-in-IP tunnels. But I don't mind either to discuss both concurrently. After all, a particular p2p H-LSP should be able to concatenate an MPLS-LSP followed by a non-MPLS-LSP followed by an MPLS-LSP again, followed by.... Heinrich -----Ursprüngliche Nachricht----- Von: Eric Rosen [mailto:erosen@cisco.com] Gesendet: Mittwoch, 14. August 2002 18:48 An: mpls@UU.NET; Loa Andersson; George Swallow Betreff: Encapsulating MPLS in IP or GRE In support of work going on in PPVPN and PWE3, it is desirable to have some standards for encapsulating MPLS packets inside IP packets. The draft whose announcement I attach below is the result of merging two previous (but expired) documents, draft-worster-mpls-in-ip, and draft-rekhter-mpls-over- gre. I would like to present this for the consideration of the MPLS WG as a standards track document. |
|