Personal T. Melsen Internet-Draft S. Blake Expires: August 2004 Ericsson February 2004 MAC Forced Forwarding: An ARP proxy method for ensuring traffic separation between hosts sharing an Ethernet access network draft-melsen-mac-forced-fwd-01.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of 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. This Internet-Draft will expire on August 13, 2004. Copyright Notice Copyright (C) The Internet Society (2004). All Rights Reserved. Abstract This document describes a mechanism to ensure layer-2 separation of LAN stations accessing an IPv4 gateway over a shared Ethernet segment. The mechanism - called "MAC Forced Forwarding" - implements an ARP proxy function that prohibits MAC address resolution between hosts located within the same IP subnet but at different customer premises, and in effect directs all upstream traffic to the IP gateway. The IP gateway provides IP-layer connectivity between these same hosts. Melsen & Blake Expires August 13, 2004 [Page 1] Internet-Draft MAC Forced Forwarding February 2004 Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1 Using Ethernet as an Access Network Technology . . . . . . . . 4 1.2 Solution Characteristics . . . . . . . . . . . . . . . . . . . 5 2. Conventions used in this document . . . . . . . . . . . . . . 5 3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 4. Solution Aspects . . . . . . . . . . . . . . . . . . . . . . . 6 4.1 Obtaining the IP and MAC addresses of the Access Router . . . 6 4.2 Responding to ARP Requests . . . . . . . . . . . . . . . . . . 6 4.3 Filtering Upstream Traffic . . . . . . . . . . . . . . . . . . 7 5. Access Router Considerations . . . . . . . . . . . . . . . . . 7 6. Resiliency Considerations . . . . . . . . . . . . . . . . . . 7 7. Multicast Considerations . . . . . . . . . . . . . . . . . . . 8 8. IPv6 Considerations . . . . . . . . . . . . . . . . . . . . . 8 9. Security Considerations . . . . . . . . . . . . . . . . . . . 9 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 9 References . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . 10 Intellectual Property and Copyright Statements . . . . . . . . 11 Melsen & Blake Expires August 13, 2004 [Page 2] Internet-Draft MAC Forced Forwarding February 2004 1. Introduction The main purpose of a remote access network is to provide connectivity between customer hosts and service provider access routers (AR), typically offering access to the Internet and other IP networks and/or IP-based applications. A remote access network may be decomposed into a subscriber line part and an aggregation network part. The subscriber line - often referred to as "the first mile" - is characterized by an individual physical connection to each customer premise. The aggregation network - "the second mile" - performs aggregation and concentration of customer traffic. The subscriber line and the aggregation network are interconnected by an Access Node (AN). Thus, the AN constitutes the border between individual subscriber lines and the common aggregation network. This is illustrated in the following figure. Access Aggregation Access Subscriber Customer Routers Network Nodes Lines Premise Networks +----+ | --+ AR +-----------| +----+ +----+ | | +-----------------[] |--------+ AN | | | +-----------------[] | +----+ | | +----+ | | +-----------------[] |--------+ AN | | | +-----------------[] | +----+ | | +----+ | | +-----------------[] |--------+ AN | +----+ | | +-----------------[] --+ AR +-----------| +----+ +----+ | It is often strongly desired that all traffic to and from customer hosts located at different premises (i.e., accessed via different subscriber lines, or via different access networks) be forwarded via an AR. This enables the access network service provider to use the Melsen & Blake Expires August 13, 2004 [Page 3] Internet-Draft MAC Forced Forwarding February 2004 AR(s) to perform security filtering, policing, and accounting of all customer traffic. This implies that within the access network, layer-2 traffic paths should not exist that circumvent an AR. In ATM-based access networks, the separation of individual customer hosts' traffic is an intrinsic feature achieved by the use of ATM permanent virtual connections (PVCs) between the customers' access device (e.g., DSL modem) and the AR (typically co-located/integrated with access control functionality in a broadband access server (BRAS)). In this case, the Access Node is an ATM-based DSLAM. This document, however, targets traffic separation for Ethernet-based access networks, i.e., where techniques other than ATM PVCs are deployed to ensure the desired separation of traffic belonging to individual customer hosts. 1.1 Using Ethernet as an Access Network Technology A major aspect of using Ethernet as an access technology is that traffic pertaining to different customer hosts is conveyed over a shared broadcast network. To avoid IP routing in the access network, the Ethernet aggregation network is bridged via Access Nodes to individual Ethernet networks at the customers' premises. In this architecture there is direct visibility between Ethernet stations (hosts) located at different customers' premises due to the nature of Ethernet. Specifically, hosts located within the same IP subnet will have this functionality. This not only violates the requirement to send all traffic via the AR, it also introduces security issues, as malicious end-users can attack hosts at other customers' premises directly at the Ethernet layer. Existing standardized solutions may be deployed to prevent layer-2 visibility between stations: o PPP over Ethernet. The use of PPPoE creates individual tunnels between hosts and one or more Access Concentrators (AC) over a bridged Ethernet topology. Traffic always flows between an AC and hosts, never between hosts. The Access Node can enforce that upstream traffic will only go to the AC initially selected by the host. o VLAN per customer. Individual traffic streams can be separated in different VLANs between the AN and the AR. Both solutions provide layer-2 isolation between customer hosts, but still they are not considered optimal for broadband remote access networks, because: o PPPoE does not support efficient multicast, one of the major advantages of using Ethernet as an access technology (instead of ATM). Melsen & Blake Expires August 13, 2004 [Page 4] Internet-Draft MAC Forced Forwarding February 2004 o Using VLANs to separate individual customer hosts is not appealing, since that is regarded as requiring cumbersome provisioning. Furthermore, the basic limit of a maximum of 4096 VLANs per-Ethernet network does not present a scalable solution. This limit is removed by deploying VLAN stacking techniques within the access network, but this solution only adds to the provisioning complexity. 1.2 Solution Characteristics The solution proposed in this document has the following main characteristics: 1. Traffic between individual customer hosts is isolated over the Ethernet access network. Traffic always flows between customer hosts and the AR, and never directly between customer hosts at different premises. 2. IP addresses are assigned to customer hosts in an efficient manner. Specifically, allocating individual IP subnets to each customer network is NOT a requirement for this solution to function. See RFC 3069 [3] for a discussion on why this requirement is relevant. 3. IP over Ethernet is used as the access protocol to ensure efficient multicast support (see Section 7). 4. VLANs are NOT used to separate traffic pertaining to individual customer hosts, due to scalability and provisioning issues. 5. The solution works for both dynamically assigned IP addresses (via DHCP) and statically assigned IP addresses. 2. Conventions used in this document In this document, the key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" are to be interpreted as described in BCP 14, RFC 2119 [2] and indicate requirement levels for compliant implementations. 3. Terminology Ethernet Access Node (EAN) The entity interconnecting individual subscriber lines and the shared aggregation network. e.g., for xDSL access, the EAN is an Ethernet-centric DSLAM (Digital Subscriber Line Access Multiplexer). The EAN is a special type of filtering bridge that does not forward Ethernet broadcast and multicast frames originating on a subscriber line to other subscriber lines, but either discards them or forwards them to an AR. The EAN also discards unicast Ethernet frames originating on a subscriber line and not addressed to an AR. Melsen & Blake Expires August 13, 2004 [Page 5] Internet-Draft MAC Forced Forwarding February 2004 4. Solution Aspects The basic property of the solution is that the EAN ensures that upstream traffic is always sent to the designated AR, even if the IP traffic goes between customer hosts located in the same IP subnet. The solution has three major aspects: 1. Initially, the EAN obtains the IP and MAC address of the target AR. 2. The EAN replies with this MAC address to any upstream ARP request from customer hosts. 3. The EAN filters out any upstream traffic to MAC addresses other than the target AR. These aspects are discussed in the following sections. 4.1 Obtaining the IP and MAC addresses of the Access Router The AR is typically the default gateway of the host. The EAN may learn the IP address of the AR in one of two ways, depending on the host IP address assignment method. If each host uses DHCP, the AR IP address is dynamically learned by snooping the DHCP reply to a host. Otherwise, the AR IP address is pre-provisioned by the network operator. In both cases, the EAN will need to determine the corresponding MAC address, using ARP. This can be done immediately after the IP address is learned, or when the MAC address is first required. An access network may contain multiple ARs, and different hosts may be assigned different ARs. This implies that the EAN MUST register the assigned AR address on a per-host basis. 4.2 Responding to ARP Requests If all customer networks were assigned individual IP subnets, all upstream traffic would normally go to an AR (typically the default gateway), and the EAN could validate all upstream traffic by checking that the destination MAC address matched the AR. However, to comply with requirement 2 of Section 1.2, residential customer networks are not assigned individual IP subnets. In other words, several hosts located at different premises are within the same IP subnet. Consequently, if a host wishes to communicate with a host at another premise, an ARP is issued to obtain that host's corresponding MAC address. This ARP request is intercepted by the EAN's ARP proxy, and responded to with an ARP reply, indicating the AR MAC address as the requested layer-2 destination. In this way, the Melsen & Blake Expires August 13, 2004 [Page 6] Internet-Draft MAC Forced Forwarding February 2004 ARP table of the requesting host will register the AR MAC address as the layer-2 destination for any host within that IP subnet. An exception is made when a host is ARPing for another host located within the same premise. If this ARP request reaches the EAN, it is discarded, because it is assumed to be answered directly by a host locally within the premise. 4.3 Filtering Upstream Traffic Since the EAN's ARP proxy will always reply with the MAC address of the AR, the requesting host will never learn MAC addresses of hosts located at other premises. However, malicious customers or malfunctioning hosts may still try to send traffic using other destination MAC addresses. This traffic MUST be discarded by the EAN. 5. Access Router Considerations Traffic between hosts that belong to the same IP subnet but are located at different premises will always be forwarded via an AR. In this case, the AR will forward the traffic to the originating network, i.e., on the same interface from where it was received. This normally results in an ICMP redirect message, RFC 792 [4], being sent to the originating host. To prevent this behavior, the ICMP redirect function for aggregation network interfaces MUST be disabled in the AR. 6. Resiliency Considerations The operation of MAC Forced Forwarding does not interfere with or delay IP connectivity recovery in the event of a sustained AR failure when DHCP is used as the IP address allocation mechanism, or when two or more ARs implement VRRP [5]. MAC Forced Forwarding is a stateful protocol. If static IP address assignment is used in the access network, then the EAN state database can be quickly recovered in the event of a transient EAN failure. Otherwise, transient failure of an EAN can lead to sustained loss of connectivity, since the DHCP and ARP messages that are snooped to construct the EAN state database are usually infrequent, and a transient failure may not be detected by either the AR(s) or the customer premise hosts. EANs used in access networks using dynamic IP address assignment MUST employ resilient storage of their state database to permit timely restoration of connectivity in the event of a transient EAN failure. Melsen & Blake Expires August 13, 2004 [Page 7] Internet-Draft MAC Forced Forwarding February 2004 7. Multicast Considerations Multicast traffic delivery for streams originating upstream from the access network and delivered to one or more customer premise hosts in an access network is supported in a scalable manner by virtue of Ethernet's native multicast capability. Efficiency can be enhanced if the EAN behaves as an IGMP snooping bridge; e.g., if it snoops on IGMP Membership Report and Leave Group messages originating on subscriber lines, to prune the set of subscriber lines on which to forward multicast packets. Support for customer-originated multicast streams is more complicated. The access network must avoid reflecting a multicast packet received on a subscriber line back out onto that line (to suppress duplicate packet reception on hosts at that customer premise). The EAN could bridge multicast frames onto (listening) subscriber lines excluding the originating one, but this would violate the requirement that all customer-originated traffic be forwarded via an AR. Alternatively, the multicast packets could be sent upstream to the appropriate AR, which would forward them back out onto the same subnet. This would require non-standard behavior on the part of the AR, and would require the EAN to filter downstream multicast traffic based on source IP address (not delivering multicast packets to the subscriber line serving the originating host). 8. IPv6 Considerations MAC Forced Forwarding is not directly applicable for IPv6 access networks, for the following reasons: 1. IPv6 access networks do not require the same efficiency of address allocation as IPv4 access networks. It is expected that customer premise networks will be allocated unique network prefixes (e.g., /48) accommodating large numbers of customer subnets and hosts. 2. IPv6 nodes do not use ARP, but instead use the Neighbor Discovery protocol [6] for layer-2 address resolution. 3. IPv6 nodes do not necessarily use DHCP to obtain IP addresses and IP gateway information, but may instead use the Stateless Address Autoconfiguration protocol [7]. Furthermore, there is no practical deployment experience using a MAC Forced Forwarding-type approach in an IPv6 access network. Some principles of MAC Address Forwarding may be applicable in an IPv6 access network design and merit further study. Melsen & Blake Expires August 13, 2004 [Page 8] Internet-Draft MAC Forced Forwarding February 2004 9. Security Considerations MAC Forced Forwarding is by its nature a security function, ensuring layer-2 isolation of customer hosts sharing a broadcast access medium. In that sense it provides security equivalent to alternative PVC-based solutions. A MAC Forced Forwarding implementation MUST ensure that only authentic DHCP replies are used in the dynamic discovery of AR addresses. One way to accomplish this is to reject any upstream DHCP replies, i.e., replies originated on a subscriber line. 10. Acknowledgements The authors would like to thank Ulf Jonsson, Thomas Narten, James Carlson, Rolf Engstrand, and Johan Kolhi for their helpful comments. References [1] Bradner, S., "The Internet Standards Process -- Revision 3", BCP 9, RFC 2026, October 1996. [2] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [3] McPherson, D. and B. Dykes, "VLAN Aggregation for Efficient IP Address Allocation", RFC 3069, February 2001. [4] Postel, J., "Internet Control Message Protocol", STD 5, RFC 792, September 1981. [5] Knight, S., Weaver, D., Whipple, D., Hinden, R., Mitzel, D., Hunt, P., Higginson, P., Shand, M. and A. Lindem, "Virtual Router Redundancy Protocol", RFC 2338, April 1998. [6] Narten, T., Nordmark, E. and W. Simpson, "Neighbor Discovery for IP Version 6 (IPv6)", RFC 2461, December 1998. [7] Thomson, S. and T. Narten, "IPv6 Stateless Address Autoconfiguration", RFC 2462, December 1998. Melsen & Blake Expires August 13, 2004 [Page 9] Internet-Draft MAC Forced Forwarding February 2004 Authors' Addresses Torben Melsen Ericsson Faelledvej Struer DK-7600 Denmark EMail: Torben.Melsen@ericsson.com Steven Blake Ericsson IP Infrastructure 920 Main Campus Drive Suite 500 Raleigh, NC 27606 USA Phone: +1 919 472-9913 EMail: steven.blake@ericsson.com Melsen & Blake Expires August 13, 2004 [Page 10] Internet-Draft MAC Forced Forwarding February 2004 Intellectual Property Statement The IETF takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any effort to identify any such rights. Information on the IETF's procedures with respect to rights in standards-track and standards-related documentation can be found in BCP-11. Copies of claims of rights made available for publication and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementors or users of this specification can be obtained from the IETF Secretariat. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights which may cover technology that may be required to practice this standard. Please address the information to the IETF Executive Director. Full Copyright Statement Copyright (C) The Internet Society (2004). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assignees. This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION Melsen & Blake Expires August 13, 2004 [Page 11] Internet-Draft MAC Forced Forwarding February 2004 HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Acknowledgment Funding for the RFC Editor function is currently provided by the Internet Society. Melsen & Blake Expires August 13, 2004 [Page 12]