INTERNET-DRAFT E. Nordmark October 27, 2003 Sun Microsystems, Inc. Obsoletes: 2893 R. E. Gilligan Intransa, Inc. Basic Transition Mechanisms for IPv6 Hosts and Routers Status of this Memo This document is an Internet-Draft and is subject to 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 draft expires on April 27, 2004. Abstract This document specifies IPv4 compatibility mechanisms that can be implemented by IPv6 hosts and routers. These mechanisms include providing complete implementations of both versions of the Internet Protocol (IPv4 and IPv6), and tunneling IPv6 packets over IPv4 routing infrastructures. They are designed to allow IPv6 nodes to maintain complete compatibility with IPv4, which should greatly simplify the deployment of IPv6 in the Internet, and facilitate the eventual transition of the entire Internet to IPv6. This document obsoletes RFC 2893. [Page 1] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 Contents Status of this Memo.......................................... 1 1. Introduction............................................. 3 1.1. Terminology......................................... 3 2. Dual IP Layer Operation.................................. 5 2.1. Address Configuration............................... 6 2.2. DNS................................................. 6 2.3. Advertising Addresses in the DNS.................... 7 3. Configured Tunneling Mechanisms.......................... 8 3.1. Encapsulation....................................... 10 3.2. Tunnel MTU and Fragmentation........................ 10 3.2.1. Static Tunnel MTU.............................. 11 3.2.2. Dynamic Tunnel MTU............................. 11 3.3. Hop Limit........................................... 13 3.4. Handling IPv4 ICMP errors........................... 13 3.5. IPv4 Header Construction............................ 14 3.6. Decapsulation....................................... 16 3.7. Link-Local Addresses................................ 18 3.8. Neighbor Discovery over Tunnels..................... 18 4. Threat Related to Source Address Spoofing................ 19 5. Security Considerations.................................. 20 6. Acknowledgments.......................................... 21 7. References............................................... 21 7.1. Normative References................................ 21 7.2. Non-normative References............................ 21 8. Authors' Addresses....................................... 23 9. Changes from RFC 2893.................................... 23 9.1. Changes from draft-ietf-v6ops-mech-v2-00............ 25 10. Open Issues............................................. 26 [Page 2] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 1. Introduction The key to a successful IPv6 transition is compatibility with the large installed base of IPv4 hosts and routers. Maintaining compatibility with IPv4 while deploying IPv6 will streamline the task of transitioning the Internet to IPv6. This specification defines a set of mechanisms that IPv6 hosts and routers may implement in order to be compatible with IPv4 hosts and routers. The mechanisms in this document are designed to be employed by IPv6 hosts and routers that need to interoperate with IPv4 hosts and utilize IPv4 routing infrastructures. We expect that most nodes in the Internet will need such compatibility for a long time to come, and perhaps even indefinitely. The mechanisms specified here include: - Dual IP layer (also known as Dual Stack): A technique for providing complete support for both Internet protocols -- IPv4 and IPv6 -- in hosts and routers. - Configured tunneling of IPv6 over IPv4: Point-to-point tunnels made by encapsulating IPv6 packets within IPv4 headers to carry them over IPv4 routing infrastructures. The mechanisms defined here are intended to be the core of a "transition toolbox" -- a growing collection of techniques which implementations and users may employ to ease the transition. The tools may be used as needed. Implementations and sites decide which techniques are appropriate to their specific needs. This document defines the basic set of transition mechanisms, but these are not the only tools available. Additional transition and compatibility mechanisms are specified in other documents. 1.1. Terminology The following terms are used in this document: Types of Nodes IPv4-only node: A host or router that implements only IPv4. An IPv4- only node does not understand IPv6. The installed base of IPv4 hosts and routers existing before the transition [Page 3] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 begins are IPv4-only nodes. IPv6/IPv4 node: A host or router that implements both IPv4 and IPv6. IPv6-only node: A host or router that implements IPv6, and does not implement IPv4. The operation of IPv6-only nodes is not addressed here. IPv6 node: Any host or router that implements IPv6. IPv6/IPv4 and IPv6-only nodes are both IPv6 nodes. IPv4 node: Any host or router that implements IPv4. IPv6/IPv4 and IPv4-only nodes are both IPv4 nodes. Types of IPv6 Addresses IPv4-compatible IPv6 address: An IPv6 address bearing the high-order 96-bit prefix 0:0:0:0:0:0, and an IPv4 address in the low-order 32- bits. IPv4-compatible addresses are no longer used by this specification, thus this definition is preserved in the specification merely to clarify their non-use. Techniques Used in the Transition IPv6-over-IPv4 tunneling: The technique of encapsulating IPv6 packets within IPv4 so that they can be carried across IPv4 routing infrastructures. Configured tunneling: IPv6-over-IPv4 tunneling where the IPv4 tunnel endpoint address is determined by configuration information on the encapsulator. The tunnels can be either unidirectional or bidirectional. Bidirectional [Page 4] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 configured tunnels behave as virtual point-to-point links. Other transition mechanisms, including other tunneling mechanisms, are outside the scope of this document. The keywords MUST, MUST NOT, REQUIRED, SHALL, SHALL NOT, SHOULD, SHOULD NOT, RECOMMENDED, MAY, and OPTIONAL, when they appear in this document, are to be interpreted as described in [RFC2119]. 2. Dual IP Layer Operation The most straightforward way for IPv6 nodes to remain compatible with IPv4-only nodes is by providing a complete IPv4 implementation. IPv6 nodes that provide a complete IPv4 and IPv6 implementations are called "IPv6/IPv4 nodes." IPv6/IPv4 nodes have the ability to send and receive both IPv4 and IPv6 packets. They can directly interoperate with IPv4 nodes using IPv4 packets, and also directly interoperate with IPv6 nodes using IPv6 packets. Even though a node may be equipped to support both protocols, one or the other stack may be disabled for operational reasons. Here we use a rather loose notion of "stack". A stack being enabled has IP addresses assigned etc, but whether or not any particular application is available on the stacks is explicitly not defined. Thus IPv6/IPv4 nodes may be operated in one of three modes: - With their IPv4 stack enabled and their IPv6 stack disabled. - With their IPv6 stack enabled and their IPv4 stack disabled. - With both stacks enabled. IPv6/IPv4 nodes with their IPv6 stack disabled will operate like IPv4-only nodes. Similarly, IPv6/IPv4 nodes with their IPv4 stacks disabled will operate like IPv6-only nodes. IPv6/IPv4 nodes MAY provide a configuration switch to disable either their IPv4 or IPv6 stack. The configured tunneling technique, which is described in section 3, may or may not be used in addition to the dual IP layer operation. An IPv6/IPv4 node MAY support configured tunneling. [Page 5] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 2.1. Address Configuration Because they support both protocols, IPv6/IPv4 nodes may be configured with both IPv4 and IPv6 addresses. IPv6/IPv4 nodes use IPv4 mechanisms (e.g., DHCP) to acquire their IPv4 addresses, and IPv6 protocol mechanisms (e.g., stateless address autoconfiguration and/or DHCPv6) to acquire their IPv6 addresses. 2.2. DNS The Domain Naming System (DNS) is used in both IPv4 and IPv6 to map between hostnames and IP addresses. A new resource record type named "AAAA" has been defined for IPv6 addresses [RFC3596]. Since IPv6/IPv4 nodes must be able to interoperate directly with both IPv4 and IPv6 nodes, they must provide resolver libraries capable of dealing with IPv4 "A" records as well as IPv6 "AAAA" records. Note that the lookup of A versus AAAA records is independent of whether the DNS packets are carried in IPv4 or IPv6 packets, and that there is no assumption that the DNS server know the IPv4/IPv6 capabilities of the requesting node. DNS resolver libraries on IPv6/IPv4 nodes MUST be capable of handling both AAAA and A records. However, when a query locates an AAAA record holding an IPv6 address, and an A record holding an IPv4 address, the resolver library MAY filter or order the results returned to the application in order to influence the version of IP packets used to communicate with that node. In terms of filtering, the resolver library has three alternatives: - Return only the IPv6 address(es) to the application. - Return only the IPv4 address(es) to the application. - Return both types of addresses to the application. If it returns only the IPv6 address(es), the application will communicate with the node using IPv6. If it returns only the IPv4 address(es), the application will communicate with the node using IPv4. If it returns both types of addresses, the application will have the choice which address to use, and thus which IP protocol to employ. If it returns both, the resolver MAY elect to order the addresses -- IPv6 first, or IPv4 first. Since most applications try the addresses in the order they are returned by the resolver, this can affect the IP version "preference" of applications. [Page 6] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 A resolver library performing filtering or ordering of addresses might also want to take into account external factors such as, whether IPv6 interfaces have been configured on the node. The decision to filter or order DNS results is implementation specific. IPv6/IPv4 nodes MAY provide policy configuration to control filtering or ordering of addresses returned by the resolver, or leave the decision entirely up to the application. An implementation MUST allow the application to control whether or not such filtering takes place. More details on the relative preferences of IPv4 and IPv6 addresses are specified in the default address selection document [RFC3484]. 2.3. Advertising Addresses in the DNS There are some constraint placed on the use of the DNS during transition. The constraints allow nodes to prefer either IPv6 or IPv4 addresses when both types of addresses are returned by the DNS. Most of these are obvious but are stated here for completeness, especially given that there is some existing practise in IPv4 to advertise unreachable IPv4 addresses in the DNS. The recommendation is that AAAA records for a node should not be added to the DNS until all of these are true: 1) The address is assigned to the interface on the node. 2) The address is configured on the interface. 3) The interface is on a link which is connected to the IPv6 infrastructure. If an IPv6 node is isolated from an IPv6 perspective (e.g., it is not connected to the 6bone to take a concrete example) constraint #3 would mean that it should not have an address in the DNS. This works great when other dual stack nodes try to contact the isolated dual stack node. There is no IPv6 address in the DNS thus the peer doesn't even try communicating using IPv6 but goes directly to IPv4 (we are assuming both nodes have A records in the DNS.) However, this does not work well when the isolated node is trying to establish communication. Even though it does not have an IPv6 address in the DNS it will find AAAA records in the DNS for the peer. [Page 7] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 Since the isolated node has IPv6 addresses assigned to at least one interface it will try to communicate using IPv6. If it has no IPv6 route to the 6bone (e.g., because the local router was upgraded to advertise IPv6 addresses using Neighbor Discovery but that router doesn't have any IPv6 routes) this communication will fail. Typically this means a few minutes of delay as TCP times out. The TCP specification [RFC1122] says that ICMP unreachable messages could be due to routing transients thus they should not immediately terminate the TCP connection. This means that the normal TCP timeout of a few minutes apply. Once TCP times out the application will hopefully try the IPv4 addresses based on the A records in the DNS, but this will be painfully slow. A possible implication of the recommendations above is that, if one enables IPv6 on a node on a link without IPv6 infrastructure, and choose to add AAAA records to the DNS for that node, then external IPv6 nodes that might see these AAAA records will possibly try to reach that node using IPv6 and suffer delays or communication failure due to unreachability. (A delay is incurred if the application correctly falls back to using IPv4 if it can not establish communication using IPv6 addresses. If this fallback is not done the application would fail to communicate in this case.) Thus it is suggested that either the recommendations be followed, or care be taken to only do so with nodes that will not be impacted by external accessing delays and/or communication failure. In the future, when a node discontinues its use of IPv4, analogous constraints apply with respect to the node's A records in the DNS; the removal of the A records should be tied to when the node can no longer be reached using IPv4. 3. Configured Tunneling Mechanisms In most deployment scenarios, the IPv6 routing infrastructure will be built up over time. While the IPv6 infrastructure is being deployed, the existing IPv4 routing infrastructure can remain functional, and can be used to carry IPv6 traffic. Tunneling provides a way to utilize an existing IPv4 routing infrastructure to carry IPv6 traffic. IPv6/IPv4 hosts and routers can tunnel IPv6 datagrams over regions of IPv4 routing topology by encapsulating them within IPv4 packets. Tunneling can be used in a variety of ways: - Router-to-Router. IPv6/IPv4 routers interconnected by an IPv4 [Page 8] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 infrastructure can tunnel IPv6 packets between themselves. In this case, the tunnel spans one segment of the end-to-end path that the IPv6 packet takes. - Host-to-Router. IPv6/IPv4 hosts can tunnel IPv6 packets to an intermediary IPv6/IPv4 router that is reachable via an IPv4 infrastructure. This type of tunnel spans the first segment of the packet's end-to-end path. - Host-to-Host. IPv6/IPv4 hosts that are interconnected by an IPv4 infrastructure can tunnel IPv6 packets between themselves. In this case, the tunnel spans the entire end-to-end path that the packet takes. - Router-to-Host. IPv6/IPv4 routers can tunnel IPv6 packets to their final destination IPv6/IPv4 host. This tunnel spans only the last segment of the end-to-end path. Configured tunneling can be used in all of the above cases, but is most likely to be used router-to-router due to the need to explicitly configure the tunneling endpoints. The underlying mechanisms for tunneling are: - The entry node of the tunnel (the encapsulator) creates an encapsulating IPv4 header and transmits the encapsulated packet. - The exit node of the tunnel (the decapsulator) receives the encapsulated packet, reassembles the packet if needed, removes the IPv4 header, updates the IPv6 header, and processes the received IPv6 packet. - The encapsulator MAY need to maintain soft state information for each tunnel recording such parameters as the MTU of the tunnel in order to process IPv6 packets forwarded into the tunnel. In cases where the number of tunnels that any one host or router is using is large, it is helpful to observe that this state information can be cached and discarded when not in use. In configured tunneling, the tunnel endpoint address is determined from configuration information in the encapsulator. For each tunnel, the encapsulator must store the tunnel endpoint address. When an IPv6 packet is transmitted over a tunnel, the tunnel endpoint address configured for that tunnel is used as the destination address for the encapsulating IPv4 header. The determination of which packets to tunnel is usually made by routing information on the encapsulator. This is usually done via a [Page 9] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 routing table, which directs packets based on their destination address using the prefix mask and match technique. 3.1. Encapsulation The encapsulation of an IPv6 datagram in IPv4 is shown below: +-------------+ | IPv4 | | Header | +-------------+ +-------------+ | IPv6 | | IPv6 | | Header | | Header | +-------------+ +-------------+ | Transport | | Transport | | Layer | ===> | Layer | | Header | | Header | +-------------+ +-------------+ | | | | ~ Data ~ ~ Data ~ | | | | +-------------+ +-------------+ Encapsulating IPv6 in IPv4 In addition to adding an IPv4 header, the encapsulator also has to handle some more complex issues: - Determine when to fragment and when to report an ICMP "packet too big" error back to the source. - How to reflect IPv4 ICMP errors from routers along the tunnel path back to the source as IPv6 ICMP errors. Those issues are discussed in the following sections. 3.2. Tunnel MTU and Fragmentation Naively the encapsulator could view encapsulation as IPv6 using IPv4 as a link layer with a very large MTU (65535-20 bytes to be exact; 20 bytes "extra" are needed for the encapsulating IPv4 header). The encapsulator would need only to report IPv6 ICMP "packet too big" [Page 10] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 errors back to the source for packets that exceed this MTU. However, such a scheme would be inefficient for two reasons and therefor MUST NOT be used: 1) It would result in more fragmentation than needed. IPv4 layer fragmentation SHOULD be avoided due to the performance problems caused by the loss unit being smaller than the retransmission unit [KM97]. 2) Any IPv4 fragmentation occurring inside the tunnel, i.e. between the encapsulator and the decapsulator, would have to be reassembled at the tunnel endpoint. For tunnels that terminate at a router, this would require additional memory to reassemble the IPv4 fragments into a complete IPv6 packet before that packet could be forwarded onward. Hence, the encapsulator MUST NOT treat the tunnel as an interface with an MTU of 64 kilobytes, but instead either use the fixed static MTU below, or use OPTIONAL dynamic MTU determination based on the IPv4 path MTU to the tunnel endpoint. 3.2.1. Static Tunnel MTU A node using a static tunnel MTU MUST limit the size of the IPv6 packets it tunnels to 1280 bytes i.e., treat the tunnel interface as having a fixed interface MTU of 1280 bytes. An implementation MAY have a configuration knob which can be used to set a larger value of the tunnel MTU than 1280 bytes, but if so the default MUST be 1280 bytes. A larger fixed MTU should not be configured unless it has been administratively ensured that the decapsulator can reassemble packets of that size. Care should be taken when manually configuring large tunnel MTUs to only do so when the MTU of the IPv4 path to the tunnel endpoint is large to avoid causing excessive fragmentation. When using the static tunnel MTU the Don't Fragment bit MUST NOT be set in the encapsulating IPv4 header. As a result the encapsulator should not receive any ICMPv4 "packet too big" message as a result of the packets it has encapsulated. 3.2.2. Dynamic Tunnel MTU The dynamic MTU determination is OPTIONAL. However, if it is implemented, it SHOULD have the behavior described in this document. [Page 11] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 The fragmentation inside the tunnel can be reduced to a minimum by having the encapsulator track the IPv4 Path MTU across the tunnel, using the IPv4 Path MTU Discovery Protocol [RFC1191] and recording the resulting path MTU. The IPv6 layer in the encapsulator can then view a tunnel as a link layer with an MTU equal to the IPv4 path MTU, minus the size of the encapsulating IPv4 header. Note that this does not eliminate IPv4 fragmentation in the case when the IPv4 path MTU would result in an IPv6 MTU less than 1280 bytes. (Any link layer used by IPv6 has to have an MTU of at least 1280 bytes [RFC2460].) In this case the IPv6 layer has to "see" a link layer with an MTU of 1280 bytes and the encapsulator has to use IPv4 fragmentation in order to forward the 1280 byte IPv6 packets. The encapsulator SHOULD employ the following algorithm to determine when to forward an IPv6 packet that is larger than the tunnel's path MTU using IPv4 fragmentation, and when to return an IPv6 ICMP "packet too big" message per [RFC1981]: if (IPv4 path MTU - 20) is less than 1280 if packet is larger than 1280 bytes Send IPv6 ICMP "packet too big" with MTU = 1280. Drop packet. else Encapsulate but do not set the Don't Fragment flag in the IPv4 header. The resulting IPv4 packet might be fragmented by the IPv4 layer on the encapsulator or by some router along the IPv4 path. endif else if packet is larger than (IPv4 path MTU - 20) Send IPv6 ICMP "packet too big" with MTU = (IPv4 path MTU - 20). Drop packet. else Encapsulate and set the Don't Fragment flag in the IPv4 header. endif endif Encapsulators that have a large number of tunnels can choose between dynamic versus static tunnel MTU on a per-tunnel endpoint basis. Note that using dynamic tunnel MTU is subject to IPv4 PMTU blackholes should the ICMPv4 "packet too big" messages be dropped by firewalls or not generated by the routers. [RFC1435, RFC2923] [Page 12] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 3.3. Hop Limit IPv6-over-IPv4 tunnels are modeled as "single-hop". That is, the IPv6 hop limit is decremented by 1 when an IPv6 packet traverses the tunnel. The single-hop model serves to hide the existence of a tunnel. The tunnel is opaque to users of the network, and is not detectable by network diagnostic tools such as traceroute. The single-hop model is implemented by having the encapsulators and decapsulators process the IPv6 hop limit field as they would if they were forwarding a packet on to any other datalink. That is, they decrement the hop limit by 1 when forwarding an IPv6 packet. (The originating node and final destination do not decrement the hop limit.) The TTL of the encapsulating IPv4 header is selected in an implementation dependent manner. The current suggested value is published in the "Assigned Numbers" RFC [RFC3232][ASSIGNED]. Implementations MAY provide a mechanism to allow the administrator to configure the IPv4 TTL such as the one specified in the IP Tunnel MIB [RFC2667]. 3.4. Handling IPv4 ICMP errors In response to encapsulated packets it has sent into the tunnel, the encapsulator might receive IPv4 ICMP error messages from IPv4 routers inside the tunnel. These packets are addressed to the encapsulator because it is the IPv4 source of the encapsulated packet. The ICMP "packet too big" error messages are handled according to IPv4 Path MTU Discovery [RFC1191] and the resulting path MTU is recorded in the IPv4 layer. The recorded path MTU is used by IPv6 to determine if an IPv6 ICMP "packet too big" error has to be generated as described in section 3.2.2 if dynamic tunnel MTU is used. The handling of other types of ICMP error messages depends on how much information is included in the "packet in error" field, which holds the encapsulated packet that caused the error. Many older IPv4 routers return only 8 bytes of data beyond the IPv4 header of the packet in error, which is not enough to include the address fields of the IPv6 header. More modern IPv4 routers are likely to return enough data beyond the IPv4 header to include the entire IPv6 header and possibly even the data beyond that. If the offending packet includes enough data, the encapsulator MAY [Page 13] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 extract the encapsulated IPv6 packet and use it to generate an IPv6 ICMP message directed back to the originating IPv6 node, as shown below: +--------------+ | IPv4 Header | | dst = encaps | | node | +--------------+ | ICMP | | Header | - - +--------------+ | IPv4 Header | | src = encaps | IPv4 | node | +--------------+ - - Packet | IPv6 | | Header | Original IPv6 in +--------------+ Packet - | Transport | Can be used to Error | Header | generate an +--------------+ IPv6 ICMP | | error message ~ Data ~ back to the source. | | - - +--------------+ - - IPv4 ICMP Error Message Returned to Encapsulating Node When receiving ICMPv4 errors as above and the errors are not "packet too big" it would be useful to log the error as an error related to the tunnel. Also, if sufficient headers are included in the error, then the originating node MAY send an ICMPv6 error of type "unreachable" with code "address unreachable" to the IPv6 source. (The "address unreachable" code is appropriate since, from the perspective of IPv6, the tunnel is a link and that code is used for link-specific errors [RFC2463]). 3.5. IPv4 Header Construction When encapsulating an IPv6 packet in an IPv4 datagram, the IPv4 header fields are set as follows: Version: [Page 14] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 4 IP Header Length in 32-bit words: 5 (There are no IPv4 options in the encapsulating header.) Type of Service: 0 unless otherwise specified. (See [RFC2983] and [RFC3168] for issues relating to the ToS byte and tunneling.) Total Length: Payload length from IPv6 header plus length of IPv6 and IPv4 headers (i.e., IPv6 payload length plus a constant 60 bytes). Identification: Generated uniquely as for any IPv4 packet transmitted by the system. Flags: Set the Don't Fragment (DF) flag as specified in section 3.2. Set the More Fragments (MF) bit as necessary if fragmenting. Fragment offset: Set as necessary if fragmenting. Time to Live: Set in implementation-specific manner. Protocol: 41 (Assigned payload type number for IPv6) Header Checksum: Calculate the checksum of the IPv4 header. Source Address: [Page 15] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 IPv4 address of outgoing interface of the encapsulator. The source address MAY alternatively be administratively specified to be a specific IPv4 address assigned to the encapsulator. This is often necessary on encapsulators with multiple IPv4 addresses to ensure that the IPv4 source address is acceptable to the decapsulator. Destination Address: IPv4 address of tunnel endpoint. Any IPv6 options are preserved in the packet (after the IPv6 header). 3.6. Decapsulation When an IPv6/IPv4 host or a router receives an IPv4 datagram that is addressed to one of its own IPv4 address, and the value of the protocol field is 41, the packet is potentially part of a tunnel and needs to be verified against the list of acceptable source addresses for tunneled packets, reassembled (if fragmented at the IPv4 level), have the IPv4 header removed and the resulting IPv6 datagram be submitted to the IPv6 layer code on the node. The decapsulator MUST verify that the tunnel source address is acceptable before further processing packets to avoid creating a hole in ingress filtering (see section 4). This check also applies to packets which are delivered to transport protocols on the decapsulator. For bidirectional configured tunnels this is done by verifying that the source address is the IPv4 address of the other end of a tunnel configured on the node. For unidirectional configured tunnels the decapsulator MUST be configured with a list of IPv4 source address prefixes that are acceptable. Such a list MUST default to having zero entries i.e., the node has to be explicitly configured to accept encapsulated packets received over unidirectional configured tunnels. Packets for which the IPv4 source address does not match SHOULD be silently dropped. A side effect of this source address verification is that the node will silently discard packets with an invalid IPv4 source address such as a multicast address, a broadcast address (255.255.255.255 and the broadcast addresses configured on the node), 0.0.0.0/8, and 127.0.0.1/8. In general, it SHOULD apply the rules for martian filtering in [RFC1812] and ingress filtering [RFC2827] on the IPv4 source address. Packets caught by these checks SHOULD be silently dropped. [Page 16] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 The decapsulator MUST be capable of reassembling an IPv4 packet that is the maximum of 1280 bytes and the largest interface MTU on the decapsulator. The 1280 byte number is a result of encapsulators that use the static MTU scheme in section 3.2.1, while encapsulators that use the dynamic scheme in section 3.2.2 can cause up to the largest interface MTU on the decapsulator to be received. (Note that it is strictly the interface MTU on the last IPv4 router *before* the decapsulator that matters, but for most links the MTU is the same between all neighbors.) This reassembly limit allows dynamic tunnel MTU determination by the encapsulator to take advantage of larger IPv4 path MTUs. An implementation MAY have a configuration knob which can be used to set a larger value of the tunnel reassembly buffers than the above number, but it MUST NOT be set below the above number. The decapsulation is shown below: +-------------+ | IPv4 | | Header | +-------------+ +-------------+ | IPv6 | | IPv6 | | Header | | Header | +-------------+ +-------------+ | Transport | | Transport | | Layer | ===> | Layer | | Header | | Header | +-------------+ +-------------+ | | | | ~ Data ~ ~ Data ~ | | | | +-------------+ +-------------+ Decapsulating IPv6 from IPv4 When decapsulating the packet, the IPv6 header is not modified. (However, see [RFC2983] and [RFC3168] for issues relating to the Type of Service byte and tunneling.) If the packet is subsequently forwarded, its hop limit is decremented by one. The decapsulator performs IPv4 reassembly before decapsulating the IPv6 packet. All IPv6 options are preserved even if the encapsulating IPv4 packet is fragmented. The encapsulating IPv4 header is discarded. The length of the IPv6 packet MUST be determined from the IPv6 payload length since the IPv4 [Page 17] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 packet might be padded (thus have a length which is larger than the IPv6 packet plus the added IPv4 header). After the decapsulation the node SHOULD silently discard a packet with an invalid IPv6 source address. This includes IPv6 multicast addresses, the IPv6 unspecified address, and the loopback address but also IPv4-compatible IPv6 source addresses where the IPv4 part of the address is an IPv4 multicast address, broadcast address (255.255.255.255 and the broadcast addresses configured on the node), 0.0.0.0/8, or 127.0.0.1/8. In general it SHOULD apply the rules for martian filtering in [RFC1812] and ingress filtering [RFC2827] on the IPv4 address embedded in IPv4-compatible source addresses. 3.7. Link-Local Addresses The configured tunnels are IPv6 interfaces (over the IPv4 "link layer") thus MUST have link-local addresses. The link-local addresses are used by routing protocols operating over the tunnels. The Interface Identifier [RFC2373] for such an Interface SHOULD be the 32-bit IPv4 address of that interface, with the bytes in the same order in which they would appear in the header of an IPv4 packet, padded at the left with zeros to a total of 64 bits. Note that the "Universal/Local" bit is zero, indicating that the Interface Identifier is not globally unique. When the host has more than one IPv4 address in use on the physical interface concerned, an administrative choice of one of these IPv4 addresses is made when forming the link-local address. The IPv6 Link-local address [RFC2373] for an IPv4 virtual interface is formed by appending the Interface Identifier, as defined above, to the prefix FE80::/64. +-------+-------+-------+-------+-------+-------+------+------+ | FE 80 00 00 00 00 00 00 | +-------+-------+-------+-------+-------+-------+------+------+ | 00 00 | 00 | 00 | IPv4 Address | +-------+-------+-------+-------+-------+-------+------+------+ 3.8. Neighbor Discovery over Tunnels For unidirectional configured tunnels most of Neighbor Discovery [RFC2461] and Stateless Address Autoconfiguration [RFC2462] does not apply; only the formation of the link-local address applies. [Page 18] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 If an implementation provides bidirectional configured tunnels it MUST at least accept and respond to the probe packets used by Neighbor Unreachability Detection (NUD) [RFC2461]. Such implementations SHOULD also send NUD probe packets to detect when the configured tunnel fails at which point the implementation can use an alternate path to reach the destination. Note that Neighbor Discovery allows that the sending of NUD probes be omitted for router to router links if the routing protocol tracks bidirectional reachability. For the purposes of Neighbor Discovery the configured tunnels specified in this document are assumed to NOT have a link-layer address, even though the link-layer (IPv4) does have address. This means that: - the sender of Neighbor Discovery packets SHOULD NOT include Source Link Layer Address options or Target Link Layer Address options on the tunnel link. - the receiver MUST, while otherwise processing the neighbor discovery packet, silently ignore the content of any Source Link Layer Address options or Target Link Layer Address options received on the tunnel link. Not using a link layer address options is consistent with how neighbor discovery is used on other point-to-point links. 4. Threat Related to Source Address Spoofing The specification above contains rules that apply ingress filtering to packets before they are decapsulated. The purpose of ingress filtering in general is specified in [RFC2827]. When IP-in-IP tunneling (independent of IP versions) is used it is important that this not be a tool to bypass any ingress filtering in use for non- tunneled packets. Thus the rules in this document are derived based on should ingress filtering be used for IPv4 and IPv6, the use of tunneling should not provide an easy way to circumvent the filtering. In this case, without specific ingress filtering checks in the decapsulator, it would be possible for an attacker to inject a packet with: - Outer IPv4 source: real IPv4 address of attacker - Outer IPv4 destination: IPv4 address of decapsulator [Page 19] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 - Inner IPv6 source: Alice which is either the decapsulator or a node close to it. - Inner IPv6 destination: Bob Even if all IPv4 routers between the attacker and the decapsulator implement IPv4 ingress filtering, and all IPv6 routers between the decapsulator and Bob implement IPv6 ingress filtering, the above spoofed packets will not be filtered out. As a result Bob will receive a packet that looks like it was sent from Alice even though the sender was some unrelated node. The solution to this is to have the decapsulator only accept encapsulated packets having explicitly configured source addresses (e.g., in the case of bidirectional tunnels, the other end of the tunnel) as specified in section 3.6. 5. Security Considerations Tunneling is not known to introduce any security holes except for the possibility to circumvent ingress filtering [RFC2827]. This specification prevent tunneling from introducing additional weaknesses when IPv4 and/or IPv6 ingress filtering is in used by requiring that decapsulators only accept packets if they have been configured to accept encapsulated packets from the IPv4 source address in the received packet. Such a check is easy to perform for bidirectional tunnels, but for uni-directional tunnels it requires a separate configuration of the IPv4 source addresses that are acceptable. An implementation of tunneling needs to be aware that while a tunnel is a link (as defined in [RFC2460]), the threat model for a tunnel might be rather different than for other links, since the tunnel potentially includes all of the Internet. The recommendations to verify that the IPv4 addresses in the encapsulated packet matches what has been configured for the tunnel, coupled with use of ingress filtering in IPv4, ameliorate some of this. In addition, an implementation must treat interfaces to different links as separate e.g. to ensure that Neighbor Discovery packets arriving on one link does not effect other links. This is especially important for tunnel links. When dropping packets due to failing to match the allowed IPv4 source addresses for a tunnel the node SHOULD NOT "acknowledge" the existence of a tunnel, otherwise this could be used to probe the [Page 20] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 acceptable tunnel endpoint addresses. For that reason the specification says that such packets SHOULD be silently discarded. 6. Acknowledgments We would like to thank the members of the IPv6 working group, the Next Generation Transition (ngtrans) working group, and the v6ops working group for their many contributions and extensive review of this document. Special thanks are due to Jim Bound, Ross Callon, Bob Hinden, Bill Manning, John Moy, Mohan Parthasarathy, Pekka Savola and Fred Templin for many helpful suggestions. 7. References 7.1. Normative References [RFC1191] Mogul, J., and S. Deering., "Path MTU Discovery", RFC 1191, November 1990. [RFC1981] McCann, J., S. Deering, and J. Mogul. "Path MTU Discovery for IP version 6", RFC 1981, August 1996. [RFC2119] S. Bradner, "Key words for use in RFCs to Indicate Requirement Levels", RFC 2119, March 1997. [RFC2460] Deering, S., and Hinden, R. "Internet Protocol, Version 6 (IPv6) Specification", RFC 2460, December 1998. [RFC2463] A. Conta, S. Deering, "Internet Control Message Protocol (ICMPv6) for the Internet Protocol Version 6 (IPv6) Specification", RFC 2463, December 1998. 7.2. Non-normative References [ASSIGNED] IANA, "Assigned numbers online database", http://www.iana.org/numbers.html [Page 21] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 [KM97] Kent, C., and J. Mogul, "Fragmentation Considered Harmful". In Proc. SIGCOMM '87 Workshop on Frontiers in Computer Communications Technology. August 1987. [RFC1122] Braden, R., "Requirements for Internet Hosts - Communication Layers", STD 3, RFC 1122, October 1989. [RFC1435] S. Knowles, "IESG Advice from Experience with Path MTU Discovery", RFC 1435, March 1993. [RFC1812] F. Baker, "Requirements for IP Version 4 Routers", RFC 1812, June 1995. [RFC2461] Narten, T., Nordmark, E., and Simpson, W. "Neighbor Discovery for IP Version 6 (IPv6)", RFC 2461, December 1998. [RFC2462] Thomson, S., and Narten, T. "IPv6 Stateless Address Autoconfiguration," RFC 2462, December 1998. [RFC2667] D. Thaler, "IP Tunnel MIB", RFC 2667, August 1999. [RFC2827] Ferguson, P., and Senie, D., "Network Ingress Filtering: Defeating Denial of Service Attacks which employ IP Source Address Spoofing", RFC 2827, May 2000. [RFC2923] K. Lahey, "TCP Problems with Path MTU Discovery", RFC 2923, September 2000. [RFC2983] D. Black, "Differentiated Services and Tunnels", RFC 2983, October 2000. [RFC3056] B. Carpenter, and K. Moore, "Connection of IPv6 Domains via IPv4 Clouds", RFC 3056, February 2001. [RFC3168] K. Ramakrishnan, S. Floyd, D. Black, "The Addition of Explicit Congestion Notification (ECN) to IP", RFC 3168, September 2001. [RFC3232] Reynolds, J., "Assigned Numbers: RFC 1700 is Replaced by an On-line Database", RFC 3232, January 2002. [RFC3484] R. Draves, "Default Address Selection for IPv6", RFC 3484, February 2003. [RFC3513] Hinden, R., and S. Deering, "IP Version 6 Addressing Architecture", RFC 3513, April 2003. [RFC3596] Thomson, S., C. Huitema, V. Ksinant, and M. Souissi, "DNS [Page 22] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 Extensions to support IP version 6", RFC 3596, October 2003. 8. Authors' Addresses Erik Nordmark Sun Microsystems Laboratories 180, avenue de l'Europe 38334 SAINT ISMIER Cedex, France Tel : +33 (0)4 76 18 88 03 Fax : +33 (0)4 76 18 88 88 Email : erik.nordmark@sun.com Robert E. Gilligan Intransa, Inc. 2870 Zanker Rd., Suite 100 San Jose, CA 95134 Tel : +1 408 678 8600 Fax : +1 408 678 8800 Email : gilligan@intransa.com, gilligan@leaf.com 9. Changes from RFC 2893 The motivation for the bulk of these changes are to simplify the document to only contain the mechanisms of wide-spread use. RFC 2893 contains a mechanism called automatic tunneling. But a much more general mechanism is specified in RFC 3056 [RFC3056] which gives each node with a (global) IPv4 address a /48 IPv6 prefix i.e., enough for a whole site. The following changes have been performed since RFC 2893: - Removed references to A6 and retained AAAA. - Removed automatic tunneling and use of IPv4-compatible addresses. - Removed default Configured Tunnel using IPv4 "Anycast Address" - Removed Source Address Selection section since this is now covered by another document ([RFC3484]). [Page 23] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 - Removed brief mention of 6over4. - Split into normative and non-normative references and other reference cleanup. - Dropped "or equal" in if (IPv4 path MTU - 20) is less than or equal to 1280 - Dropped this: However, IPv6 may be used in some environments where interoperability with IPv4 is not required. IPv6 nodes that are designed to be used in such environments need not use or even implement these mechanisms. - Clarified that the dynamic path MTU mechanism in section 3.2 is OPTIONAL but if it is implemented it should follow the rules in section 3.2. - Stated that when the dynamic PMTU is not implemented the sender MUST NOT by default send IPv6 packets larger than 1280 into the tunnel. - Stated that implementations MAY have a knob by which the MTU can be set to larger values on a tunnel by tunnel basis, but that the default MUST be 1280 and that decapsulators need to be configured to match the encapsulator's MTU. - Restated the "currently underway" language about ToS to loosely point at [RFC2983] and [RFC3168]. - Stated that IPv4 source MAY also be administratively specified. (This is especially useful on multi-interface nodes and with configured tunneling) - Fixed reference to Assigned Numbers to be to online version (with proper pointer to "Assigned Numbers is obsolete" RFC) - Clarified text about ingress filtering e.g. that it applies to packet delivered to transport protocols on the decapsulator as well as packets being forwarded by the decapsulator, and how the decapsulator's checks help when IPv4 and IPv6 ingress filtering is in place. [Page 24] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 9.1. Changes from draft-ietf-v6ops-mech-v2-00 - Clarified in section 2.2 that there is no assumption that the DNS server knows the IPv4/IPv6 capabilities of the requesting node. - Clarified in section 2.2 that a filtering resolver might want to take into account external factors e.g., whether IPv6 interfaces have been configured on the node. - Clarified in section 2.3 that part of the motivation for the section is that this is the opposite of common DNS practices in IPv4; advertising unreachable IPv4 addresses in the DNS is common. - Removed the now artificial separation in a section on "common tunneling techniques" and "configured tunneling" to make one section on "configured tunneling". - Restructured the section on tunnel MTU to make the relationship between static tunnel MTU and dynamic tunnel MTU more clear. This includes fixing the unclear language about "must be 1280 but may be configurable". - Added warning about manually configuring large tunnel MTUs causing excessive fragmentation. - Added warning about IPv4 PMTU blackholes when using dynamic MTU. - Clarified that when decapsulating the receiver must be liberal and allow for padding of the encapsulated packet. - Added example that when reflecting ICMPv4 errors as ICMPv6 errors it would be appropriate to use ICMPv6 unreachable type with code "address unreachable" since an error from inside the tunnel is in effect a link specific problem from IPv6's perspective. - Consolidated the text on ingress filtering and created a separate section on the threat related to source address spoofing through open decapsulators. - Clarified "martian" filtering as follows: 0.0.0.0 should be 0.0.0.0/8, same for 127. (per RFC1812), and elaborated that the broadcast address check includes both the 255.255.255.255 address and all the broadcast addresses of the decapsulator. [Page 25] INTERNET DRAFT Basic IPv6 Transition Mechanisms October 2003 - Clarified that packets which fail the checks (such as verifying the IPv4 source address, martian, and ingress filtering) on the decapsulator should be silently dropped. - Clarified that while source link layer address options and target link layer address options are ignored in received ND packets, the ND packets themselves are processed as normal. 10. Open Issues The document has some specific text about unidirectional configure tunnels since they are different with respect to Neighbor Discovery and ingress filtering. Does anybody implement unidirectional tunnels? Should we remove the specific text and make the explicit assumption that all configured tunnels are bidirectional? [Page 26]