Internet Engineering Task Force | F. Brockners |
Internet-Draft | S. Gundavelli |
Intended status: Standards Track | Cisco |
Expires: January 10, 2012 | S. Speicher |
Deutsche Telekom AG | |
D. Ward | |
Juniper Networks | |
July 09, 2011 |
Gateway Initiated Dual-Stack Lite Deployment
draft-ietf-softwire-gateway-init-ds-lite-05
Gateway-Initiated Dual-Stack lite (GI-DS-lite) is a variant of Dual-Stack lite (DS-lite) applicable to certain tunnel-based access architectures. GI-DS-lite extends existing access tunnels beyond the access gateway to an IPv4-IPv4 NAT using softwires with an embedded context identifier that uniquely identifies the end-system the tunneled packets belong to. The access gateway determines which portion of the traffic requires NAT using local policies and sends/receives this portion to/from this softwire.
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at http://datatracker.ietf.org/drafts/current/.
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."
This Internet-Draft will expire on January 10, 2012.
Copyright (c) 2011 IETF Trust and the persons identified as the document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
Gateway-Initiated Dual-Stack lite (GI-DS-lite) is a variant of the Dual-Stack lite (DS-lite) [I-D.ietf-softwire-dual-stack-lite], applicable to network architectures which use point to point tunnels between the access device and the access gateway. The access gateway in these models is designed to serve large numbers of access devices. Mobile architectures based on Mobile IPv6 [RFC3775], Proxy Mobile IPv6 [RFC5213], or GTP [TS29060], as well as broadband architectures based on PPP or point-to-point VLANs as defined by the Broadband Forum (see [TR59] and [TR101]) are examples for this type of architecture.
The DS-lite approach leverages IPv4-in-IPv6 tunnels (or other tunneling modes) for carrying the IPv4 traffic from the customer network to the Address Family Transition Router (AFTR). An established softwire between the AFTR and the access device is used for traffic forwarding purposes. This turns the inner IPv4 address irrelevant for traffic routing and allows sharing private IPv4 addresses [RFC1918] between customer sites within the service provider network.
Similar to DS-lite, GI-DS-lite enables the service provider to share public IPv4 addresses among different customers by combining tunneling and NAT. It allows multiple access devices behind the access gateway to share the same private IPv4 address [RFC1918]. Rather than initiating the tunnel right on the access device, GI-DS-lite logically extends the already existing access tunnels beyond the access gateway towards the Address Family Transition Router (AFTR) using a tunneling mechanism with semantics for carrying context state related to the encapsulated traffic. This approach results in supporting overlapping IPv4 addresses in the access network, requiring no changes to either the access device, or to the access architecture. Additional tunneling overhead in the access network is also omitted. If e.g., a GRE based encapsulation mechanisms is chosen, it allows the network between the access gateway and the AFTR to be either IPv4 or IPv6 and provides the operator to migrate to IPv6 in incremental steps.
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 [RFC2119].
The following abbreviations are used within this document:
The section provides an overview of Gateway Initiated DS-Lite (GI-DS-lite). Figure 1 outlines the generic deployment scenario for GI-DS-lite. This generic scenario can be mapped to multiple different access architectures, some of which are described in Appendix Appendix A.
In Figure 1, access devices (AD-1 and AD-2) are connected to the Gateway using some form of tunnel technology and the same is used for carrying IPv4 (and optionally IPv6) traffic of the access device. These access devices may also be connected to the Gateway over point-to-point links. The details on how the network delivers the IPv4 address configuration to the access devices are specific to the access architecture and are outside the scope of this document. With GI-DS-lite, Gateway and AFTR are connected by a softwire [RFC4925]. The softwire is identified by a softwire identifier (SWID). The SWID does not need to be globally unique, i.e. different SWIDs could be used to identify a softwire at the different ends of a softwire. The form of the SWID depends on the tunneling technology used for the softwire. The SWID could e.g. be the endpoints of a GRE-tunnel or a VPN-ID, see Section 6 for details. A Context-Identifier (CID) is used to multiplex flows associated with the individual access devices onto the softwire. Deployment dependent, the flows from a particular AD can be identified using either the source IP-address or an access tunnel identifier. Local policies at the Gateway determine which part of the traffic received from an access device is tunneled over the softwire to the AFTR. The combination of CID and SWID must be unique between gateway and AFTR to identify the flows associated with an AD. The CID is typically a 32-bit wide identifier and is assigned by the Gateway. It is retrieved either from a local or remote (e.g. AAA) repository. Like the SWID, the embodiment of the CID depends on the tunnel mode used and the type of the network connecting Gateway and AFTR. If, for example GRE [RFC2784] with “GRE Key and Sequence Number Extensions” [RFC2890] is used as softwire technology, the network connecting Gateway and AFTR could be either IPv4-only, IPv6-only, or a dual-stack IP network. The CID would be carried within the GRE-key field. See Section 6 for details on different softwire types supported with GI-DS-lite.
Access Device: AD-1 Context Id: CID-1 NAT Mappings: IPv4: a.b.c.d +---+ (CID-1, TCP port1 <-> +------+ access tunnel | | e.f.g.h, TCP port2) | AD-1 |=================| G | +---+ +------+ | A | | A | | T | Softwire SWID-1 | F | | E |==========================| T | IPv4: a.b.c.d | W | (e.g. IPv4-over-GRE | R | +------+ | A | over IPv4 or IPv6) +---+ | AD-2 |=================| Y | +------+ access tunnel | | (CID-2, TCP port3 <-> | | e.f.g.h, TCP port4) +---+ Access Device: AD-2 Context Id: CID-2
The AFTR combines softwire termination and IPv4-IPv4 NAT. The NAT binding of the AD's address could be assigned autonomously by the AFTR from a local address pool, configured on a per-binding basis (either by a remote control entity through a NAT control protocol or through manual configuration), or derived from the CID (e.g., the CID, in case 32-bit wide, could be mapped 1:1 to an external IPv4-address). A simple example of a translation table at the AFTR is shown in Figure 2. The choice of the appropriate translation scheme for a traffic flow can take parameters such as destination IP-address, incoming interface, etc. into account. The IP-address of the AFTR, which, depending on the transport network between the Gateway and the AFTR, will either be an IPv6 or an IPv4 address, is configured on the Gateway. A variety of methods, such as out-of-band mechanisms, or manual configuration apply.
+=====================================+======================+ | Softwire-Id/Context-Id/IPv4/Port | Public IPv4/Port | +=====================================+======================+ | SWID-1/CID-1/a.b.c.d/TCP-port1 | e.f.g.h/TCP-port2 | | | | | SWID-1/CID-2/a.b.c.d/TCP-port3 | e.f.g.h/TCP-port4 | +-------------------------------------+----------------------+
GI-DS-lite does not require a 1:1 relationship between Gateway and AFTR, but more generally applies to (M:N) scenarios, where M Gateways are connected to N AFTRs. Multiple Gateways could be served by a single AFTR. AFTRs could be dedicated to specifc groups of access-devices, groups of Gateways, or geographic regions. An AFTR could, but does not have to be co-located with a Gateway.
The following are the considerations related to the operational management of the softwire between AFTR and Gateway.
Deployment and requirements dependent, different tunnel technologies apply for the softwire connecting Gateway and AFTR. GRE encapsulation with GRE-key extensions, MPLS VPNs [RFC4364], or plain IP-in-IP encapsulation can be used. Softwire identification and Context-ID depend on the tunneling technology employed:
Figure 3 gives an overview of the different tunnel modes as they apply to different deployment scenarios. "x" indicates that a certain deployment scenario is supported. The following abbreviations are used:
+====================+==================+=======================+ | | IPv4 address | Network-type | | Softwire +----+----+----+---+----+----+------+------+ | | up | op | nm | s | v4 | v6 | v4v6 | MPLS | +====================+====+====+====+===+====+====+======+======+ | GRE with GRE-key | x | x | x | x | x | x | x | | | MPLS VPN | x | x | x | | | | | x | | IPv4/IPv6-in-MPLS | x | x | x | x | | | | x | | IPv4-in-IPv4 | x | | x | | x | | | | | IPv4-in-IPv6 | x | | x | | | x | | | | IPv4-in-IPv6 w/ FL | x | x | x | x | | x | | | +====================+====+====+====+===+====+====+======+======+
The authors would like to acknowledge the discussions on this topic with Mark Grayson, Jay Iyer, Kent Leung, Vojislav Vucetic, Flemming Andreasen, Dan Wing, Jouni Korhonen, Teemu Savolainen, Parviz Yegani, Farooq Bari, Mohamed Boucadair, Vinod Pandey, Jari Arkko, Eric Voit, Yiu L. Lee, Tina Tsou, Guo-Liang Yang, Cathy Zhou, Olaf Bonness, Paco Cortes, and Jim Guichard.
This document includes no request to IANA.
All drafts are required to have an IANA considerations section (see the update of RFC 2434 [RFC5226] for a guide). If the draft does not require IANA to do anything, the section contains an explicit statement that this is the case (as above). If there are no requirements for IANA, the section will be removed during conversion into an RFC by the RFC Editor.
All the security considerations from GTP [TS29060], Mobile IPv6 [RFC3775], Proxy Mobile IPv6 [RFC5213], and Dual-Stack lite [I-D.ietf-softwire-dual-stack-lite] apply to this specification as well.
Changes from -00 to -01
Changes from -00 to -02
Changes from -02 to -03
Changes from -03 to -04
Figure 4 shows an example call flow - linking access tunnel establishment on the Gateway with the softwire to the AFTR. This simple example assumes that traffic from the AD uses a single access tunnel and that the Gateway will use local polices to decide which portion of the traffic received over this access tunnel needs to be forwarded to the AFTR.
AD Gateway AAA/Policy AFTR | | | | |----(1)-------->| | | | (2)<-------------->| | | (3) | | | |<------(4)------------------->| | (5) | | |<---(6)-------->| | | | | | |
The section outlines deployment examples of the generic GI-DS-lite architecture described in Section 3.