Internet-Draft | IPv6 Node Requirements | October 2024 |
Chown, et al. | Expires 24 April 2025 | [Page] |
This document defines requirements for IPv6 nodes. It is expected that IPv6 will be deployed in a wide range of devices and situations. Specifying the requirements for IPv6 nodes allows IPv6 to function well and interoperate in a large number of situations and deployments.¶
This document obsoletes RFC 8504, and in turn RFC 6434 and its predecessor, RFC 4294.¶
This note is to be removed before publishing as an RFC.¶
The latest revision of this draft can be found at https://timchown.github.io/rfc8504-bis/draft-clw-6man-rfc8504-bis.html. Status information for this document may be found at https://datatracker.ietf.org/doc/draft-clw-6man-rfc8504-bis/.¶
Discussion of this document takes place on the IPv6 Maintenance Working Group mailing list (mailto:ipv6@ietf.org), which is archived at https://mailarchive.ietf.org/arch/browse/ipv6/. Subscribe at https://www.ietf.org/mailman/listinfo/ipv6/.¶
Source for this draft and an issue tracker can be found at https://github.com/timchown/rfc8504-bis.¶
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 https://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 24 April 2025.¶
Copyright (c) 2024 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 (https://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 Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
This document defines common functionality required by both IPv6 hosts and routers. Many IPv6 nodes will implement optional or additional features, but this document collects and summarizes requirements from other published Standards Track documents in one place.¶
This document tries to avoid discussion of protocol details and references RFCs for this purpose. This document is intended to be an applicability statement and to provide guidance as to which IPv6 specifications should be implemented in the general case and which specifications may be of interest to specific deployment scenarios. This document does not update any individual protocol document RFCs.¶
Although this document points to different specifications, it should be noted that in many cases, the granularity of a particular requirement will be smaller than a single specification, as many specifications define multiple, independent pieces, some of which may not be mandatory. In addition, most specifications define both client and server behavior in the same specification, while many implementations will be focused on only one of those roles.¶
This document defines a minimal level of requirement needed for a device to provide useful Internet service and considers a broad range of device types and deployment scenarios. Because of the wide range of deployment scenarios, the minimal requirements specified in this document may not be sufficient for all deployment scenarios. It is perfectly reasonable (and indeed expected) for other profiles to define additional or stricter requirements appropriate for specific usage and deployment environments. As an example, this document does not mandate that all clients support DHCP, but some deployment scenarios may deem it appropriate to make such a requirement. As another example, NIST has defined profiles for specialized requirements for IPv6 in target environments (see [USGv6]).¶
As it is not always possible for an implementer to know the exact usage of IPv6 in a node, an overriding requirement for IPv6 nodes is that they should adhere to Jon Postel's Robustness Principle: "Be conservative in what you do, be liberal in what you accept from others" [RFC0793].¶
IPv6 covers many specifications. It is intended that IPv6 will be deployed in many different situations and environments. Therefore, it is important to develop requirements for IPv6 nodes to ensure interoperability.¶
From "Internet Protocol, Version 6 (IPv6) Specification" [RFC8200], we have the following definitions:¶
IPv6 node - a device that implements IPv6. IPv6 router - a node that forwards IPv6 packets not explicitly addressed to itself. IPv6 host - any IPv6 node that is not a router.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
AH Authentication Header DAD Duplicate Address Detection ESP Encapsulating Security Payload ICMP Internet Control Message Protocol IKE Internet Key Exchange MIB Management Information Base MLD Multicast Listener Discovery MTU Maximum Transmission Unit NA Neighbor Advertisement NBMA Non-Broadcast Multi-Access ND Neighbor Discovery NS Neighbor Solicitation NUD Neighbor Unreachability Detection PPP Point-to-Point Protocol¶
An IPv6 node MUST include support for one or more IPv6 link-layer specifications. Which link-layer specifications an implementation should include will depend upon what link layers are supported by the hardware available on the system. It is possible for a conformant IPv6 node to support IPv6 on some of its interfaces and not on others.¶
As IPv6 is run over new Layer 2 technologies, it is expected that new specifications will be issued. We list here some of the Layer 2 technologies for which an IPv6 specification has been developed. It is provided for informational purposes only and may not be complete.¶
Transmission of IPv6 Packets over Ethernet Networks [RFC2464]¶
Transmission of IPv6 Packets over Frame Relay Networks Specification [RFC2590]¶
Transmission of IPv6 Packets over IEEE 1394 Networks [RFC3146]¶
Transmission of IPv6, IPv4, and Address Resolution Protocol (ARP) Packets over Fibre Channel [RFC4338]¶
Transmission of IPv6 Packets over IEEE 802.15.4 Networks [RFC4944]¶
Transmission of IPv6 via the IPv6 Convergence Sublayer over IEEE 802.16 Networks [RFC5121]¶
In addition to traditional physical link layers, it is also possible to tunnel IPv6 over other protocols. Examples include:¶
The Internet Protocol version 6 is specified in [RFC8200]. This specification MUST be supported.¶
The node MUST follow the packet transmission rules in RFC 8200.¶
All conformant IPv6 implementations MUST be capable of sending and receiving IPv6 packets; forwarding functionality MAY be supported. Nodes MUST always be able to send, receive, and process Fragment headers.¶
IPv6 nodes MUST NOT create overlapping fragments. Also, when reassembling an IPv6 datagram, if one or more of its constituent fragments is determined to be an overlapping fragment, the entire datagram (and any constituent fragments) MUST be silently discarded. See Section 4.5 of [RFC8200] for more information.¶
As recommended in Section 4.5 of [RFC8200], nodes MUST NOT generate atomic fragments, i.e., where the fragment is a whole datagram. As per [RFC6946], if a receiving node reassembling a datagram encounters an atomic fragment, it should be processed as a fully reassembled packet, and any other fragments that match this packet should be processed independently.¶
To mitigate a variety of potential attacks, nodes SHOULD avoid using predictable Fragment Identification values in Fragment headers, as discussed in [RFC7739].¶
All nodes SHOULD support the setting and use of the IPv6 Flow Label field as defined in the IPv6 Flow Label specification [RFC6437]. Forwarding nodes such as routers and load distributors MUST NOT depend only on Flow Label values being uniformly distributed. It is RECOMMENDED that source hosts support the flow label by setting the Flow Label field for all packets of a given flow to the same value chosen from an approximation to a discrete uniform distribution.¶
RFC 8200 specifies extension headers and the processing for these headers.¶
Extension headers (except for the Hop-by-Hop Options header) are not processed, inserted, or deleted by any node along a packet's delivery path, until the packet reaches the node (or each of the set of nodes, in the case of multicast) identified in the Destination Address field of the IPv6 header.¶
Any unrecognized extension headers or options MUST be processed as described in RFC 8200. Note that where Section 4 of [RFC8200] refers to the action to be taken when a Next Header value in the current header is not recognized by a node, that action applies whether the value is an unrecognized extension header or an unrecognized upper-layer protocol (ULP).¶
An IPv6 node MUST be able to process these extension headers. An exception is Routing Header type 0 (RH0), which was deprecated by [RFC5095] due to security concerns and which MUST be treated as an unrecognized routing type.¶
Further, [RFC7045] adds specific requirements for the processing of extension headers, in particular that any forwarding node along an IPv6 packet's path, which forwards the packet for any reason, SHOULD do so regardless of any extension headers that are present.¶
As per RFC 8200, when a node fragments an IPv6 datagram, it MUST include the entire IPv6 Header Chain in the first fragment. The Per-Fragment headers MUST consist of the IPv6 header plus any extension headers that MUST be processed by nodes en route to the destination, that is, all headers up to and including the Routing header if present, else the Hop-by-Hop Options header if present, else no extension headers. On reassembly, if the first fragment does not include all headers through an upper-layer header, then that fragment SHOULD be discarded and an ICMP Parameter Problem, Code 3, message SHOULD be sent to the source of the fragment, with the Pointer field set to zero. See [RFC7112] for a discussion of why oversized IPv6 Extension Header Chains are avoided.¶
Defining new IPv6 extension headers is not recommended, unless there are no existing IPv6 extension headers that can be used by specifying a new option for that IPv6 extension header. A proposal to specify a new IPv6 extension header MUST include a detailed technical explanation of why an existing IPv6 extension header can not be used for the desired new function, and in such cases, it needs to follow the format described in Section 8 of [RFC8200]. For further background reading on this topic, see [RFC6564].¶
As per RFC 8200, end hosts are expected to process all extension headers, destination options, and hop-by-hop options in a packet. Given that the only limit on the number and size of extension headers is the MTU, the processing of received packets could be considerable. It is also conceivable that a long chain of extension headers might be used as a form of denial-of-service attack. Accordingly, a host may place limits on the number and sizes of extension headers and options it is willing to process.¶
A host MAY limit the number of consecutive PAD1 options in destination options or hop-by-hop options to 7. In this case, if there are more than 7 consecutive PAD1 options present, the packet MAY be silently discarded. The rationale is that if padding of 8 or more bytes is required, then the PADN option SHOULD be used.¶
A host MAY limit the number of bytes in a PADN option to be less than 8. In such a case, if a PADN option is present that has a length greater than 7, the packet SHOULD be silently discarded. The rationale for this guideline is that the purpose of padding is for alignment and 8 bytes is the maximum alignment used in IPv6.¶
A host MAY disallow unknown options in destination options or hop-by-hop options. This SHOULD be configurable where the default is to accept unknown options and process them per [RFC8200]. If a packet with unknown options is received and the host is configured to disallow them, then the packet SHOULD be silently discarded.¶
A host MAY impose a limit on the maximum number of non-padding options allowed in the destination options and hop-by-hop extension headers. If this feature is supported, the maximum number SHOULD be configurable, and the default value SHOULD be set to 8. The limits for destination options and hop-by-hop options may be separately configurable. If a packet is received and the number of destination or hop-by-hop options exceeds the limit, then the packet SHOULD be silently discarded.¶
A host MAY impose a limit on the maximum length of Destination Options or Hop-by-Hop Options extension headers. This value SHOULD be configurable, and the default is to accept options of any length. If a packet is received and the length of the Destination or Hop-by-Hop Options extension header exceeds the length limit, then the packet SHOULD be silently discarded.¶
Neighbor Discovery is defined in [RFC4861]; the definition was updated by [RFC5942]. Neighbor Discovery MUST be supported with the noted exceptions below. RFC 4861 states:¶
Unless specified otherwise (in a document that covers operating IP over a particular link type) this document applies to all link types. However, because ND uses link-layer multicast for some of its services, it is possible that on some link types (e.g., Non‑Broadcast Multi-Access (NBMA) links), alternative protocols or mechanisms to implement those services will be specified (in the appropriate document covering the operation of IP over a particular link type). The services described in this document that are not directly dependent on multicast, such as Redirects, Next-hop determination, Neighbor Unreachability Detection, etc., are expected to be provided as specified in this document. The details of how one uses ND on NBMA links are addressed in [RFC2491].¶
Some detailed analysis of Neighbor Discovery follows:¶
Router Discovery is how hosts locate routers that reside on an attached link. Hosts MUST support Router Discovery functionality.¶
Prefix Discovery is how hosts discover the set of address prefixes that define which destinations are on-link for an attached link. Hosts MUST support Prefix Discovery.¶
Hosts MUST also implement Neighbor Unreachability Detection (NUD) for all paths between hosts and neighboring nodes. NUD is not required for paths between routers. However, all nodes MUST respond to unicast Neighbor Solicitation (NS) messages.¶
[RFC7048] discusses NUD, in particular cases where it behaves too impatiently. It states that if a node transmits more than a certain number of packets, then it SHOULD use the exponential backoff of the retransmit timer, up to a certain threshold point.¶
Hosts MUST support the sending of Router Solicitations and the receiving of Router Advertisements (RAs). The ability to understand individual RA options is dependent on supporting the functionality making use of the particular option.¶
[RFC7559] discusses packet loss resiliency for Router Solicitations and requires that nodes MUST use a specific exponential backoff algorithm for retransmission of Router Solicitations.¶
All nodes MUST support the sending and receiving of Neighbor Solicitation (NS) and Neighbor Advertisement (NA) messages. NS and NA messages are required for Duplicate Address Detection (DAD).¶
Hosts SHOULD support the processing of Redirect functionality. Routers MUST support the sending of Redirects, though not necessarily for every individual packet (e.g., due to rate limiting). Redirects are only useful on networks supporting hosts. In core networks dominated by routers, Redirects are typically disabled. The sending of Redirects SHOULD be disabled by default on routers intended to be deployed on core networks. They MAY be enabled by default on routers intended to support hosts on edge networks.¶
As specified in [RFC6980], nodes MUST NOT employ IPv6 fragmentation for sending any of the following Neighbor Discovery and SEcure Neighbor Discovery messages: Neighbor Solicitation, Neighbor Advertisement, Router Solicitation, Router Advertisement, Redirect, or Certification Path Solicitation. Nodes MUST silently ignore any of these messages on receipt if fragmented. See RFC 6980 for details and motivation.¶
"IPv6 Host-to-Router Load Sharing" [RFC4311] includes additional recommendations on how to select from a set of available routers. [RFC4311] SHOULD be supported.¶
SEND [RFC3971] and Cryptographically Generated Addresses (CGAs) [RFC3972] provide a way to secure the message exchanges of Neighbor Discovery. SEND has the potential to address certain classes of spoofing attacks, but it does not provide specific protection for threats from off-link attackers.¶
There have been relatively few implementations of SEND in common operating systems and platforms since its publication in 2005; thus, deployment experience remains very limited to date.¶
At this time, support for SEND is considered optional. Due to the complexity in deploying SEND and its heavyweight provisioning, its deployment is only likely to be considered where nodes are operating in a particularly strict security environment.¶
Router Advertisements include an 8-bit field of single-bit Router Advertisement flags. The Router Advertisement Flags Option extends the number of available flag bits by 48 bits. At the time of this writing, 6 of the original 8 single-bit flags have been assigned, while 2 remain available for future assignment. No flags have been defined that make use of the new option; thus, strictly speaking, there is no requirement to implement the option today. However, implementations that are able to pass unrecognized options to a higher-level entity that may be able to understand them (e.g., a user-level process using a "raw socket" facility) MAY take steps to handle the option in anticipation of a future usage.¶
"Path MTU Discovery for IP version 6" [RFC8201] SHOULD be supported. From [RFC8200]:¶
It is strongly recommended that IPv6 nodes implement Path MTU Discovery [RFC8201], in order to discover and take advantage of path MTUs greater than 1280 octets. However, a minimal IPv6 implementation (e.g., in a boot ROM) may simply restrict itself to sending packets no larger than 1280 octets, and omit implementation of Path MTU Discovery.¶
The rules in [RFC8200] and [RFC5722] MUST be followed for packet fragmentation and reassembly.¶
As described in RFC 8201, nodes implementing Path MTU Discovery and sending packets larger than the IPv6 minimum link MTU are susceptible to problematic connectivity if ICMPv6 messages are blocked or not transmitted. For example, this will result in connections that complete the TCP three-way handshake correctly but then hang when data is transferred. This state is referred to as a black-hole connection [RFC2923]. Path MTU Discovery relies on ICMPv6 Packet Too Big (PTB) to determine the MTU of the path (and thus these MUST NOT be filtered, as per the recommendation in [RFC4890]).¶
An alternative to Path MTU Discovery defined in RFC 8201 can be found in [RFC4821], which defines a method for Packetization Layer Path MTU Discovery (PLPMTUD) designed for use over paths where delivery of ICMPv6 messages to a host is not assured.¶
While an IPv6 link MTU can be set to 1280 bytes, it is recommended that for IPv6 UDP in particular, which includes DNS operation, the sender use a large MTU if they can, in order to avoid gratuitous fragmentation-caused packet drops.¶
ICMPv6 [RFC4443] MUST be supported. "Extended ICMP to Support Multi-Part Messages" [RFC4884] MAY be supported.¶
"Default Router Preferences and More-Specific Routes" [RFC4191] provides support for nodes attached to multiple (different) networks, each providing routers that advertise themselves as default routers via Router Advertisements. In some scenarios, one router may provide connectivity to destinations that the other router does not, and choosing the "wrong" default router can result in reachability failures. In order to resolve this scenario, IPv6 nodes MUST implement [RFC4191] and MUST implement the Type C host role defined in RFC 4191.¶
In multihomed scenarios, where a host has more than one prefix, each allocated by an upstream network that is assumed to implement BCP 38 ingress filtering, the host may have multiple routers to choose from.¶
Hosts that may be deployed in such multihomed environments SHOULD follow the guidance given in [RFC8028].¶
Nodes that need to join multicast groups MUST support MLDv2 [RFC3810]. MLD is needed by any node that is expected to receive and process multicast traffic; in particular, MLDv2 is required for support for source-specific multicast (SSM) as per [RFC4607].¶
Previous versions of this specification only required MLDv1 [RFC2710] to be implemented on all nodes. Since participation of any MLDv1-only nodes on a link require that all other nodes on the link then operate in version 1 compatibility mode, the requirement to support MLDv2 on all nodes was upgraded to a MUST. Further, SSM is now the preferred multicast distribution method, rather than Any-Source Multicast (ASM).¶
Note that Neighbor Discovery (as used on most link types -- see Section 5.4) depends on multicast and requires that nodes join Solicited Node multicast addresses.¶
An ECN-aware router sets a mark in the IP header in order to signal impending congestion, rather than dropping a packet. The receiver of the packet echoes the congestion indication to the sender, which can then reduce its transmission rate as if it detected a dropped packet.¶
Nodes SHOULD support ECN [RFC3168] by implementing an interface for the upper layer to access and by setting the ECN bits in the IP header. The benefits of using ECN are documented in [RFC8087].¶
The IPv6 Addressing Architecture [RFC4291] MUST be supported.¶
The current IPv6 Address Architecture is based on a 64-bit boundary for subnet prefixes. The reasoning behind this decision is documented in [RFC7421].¶
Implementations MUST also support the multicast flag updates documented in [RFC7371].¶
Hosts may be configured with addresses through a variety of methods, including Stateless Address Autoconfiguration (SLAAC), DHCPv6, or manual configuration.¶
[RFC7934] recommends that networks provide general-purpose end hosts with multiple global IPv6 addresses when they attach, and it describes the benefits of and the options for doing so. Routers SHOULD support [RFC7934] for assigning multiple addresses to a host. A host SHOULD support assigning multiple addresses as described in [RFC7934].¶
Nodes SHOULD support the capability to be assigned a prefix per host as documented in [RFC8273]. Such an approach can offer improved host isolation and enhanced subscriber management on shared network segments.¶
Hosts MUST support IPv6 Stateless Address Autoconfiguration. It is RECOMMENDED, as described in [RFC8064], that unless there is a specific requirement for Media Access Control (MAC) addresses to be embedded in an Interface Identifier (IID), nodes follow the procedure in [RFC7217] to generate SLAAC-based addresses, rather than use [RFC4862]. Addresses generated using the method described in [RFC7217] will be the same whenever a given device (re)appears on the same subnet (with a specific IPv6 prefix), but the IID will vary on each subnet visited.¶
Nodes that are routers MUST be able to generate link-local addresses as described in [RFC4862].¶
From RFC 4862:¶
The autoconfiguration process specified in this document applies only to hosts and not routers. Since host autoconfiguration uses information advertised by routers, routers will need to be configured by some other means. However, it is expected that routers will generate link-local addresses using the mechanism described in this document. In addition, routers are expected to successfully pass the Duplicate Address Detection procedure described in this document on all addresses prior to assigning them to an interface.¶
All nodes MUST implement Duplicate Address Detection. Quoting from Section 5.4 of [RFC4862]:¶
Duplicate Address Detection MUST be performed on all unicast addresses prior to assigning them to an interface, regardless of whether they are obtained through stateless autoconfiguration, DHCPv6, or manual configuration, with the following exceptions [noted therein].¶
"Optimistic Duplicate Address Detection (DAD) for IPv6" [RFC4429] specifies a mechanism to reduce delays associated with generating addresses via Stateless Address Autoconfiguration [RFC4862]. RFC 4429 was developed in conjunction with Mobile IPv6 in order to reduce the time needed to acquire and configure addresses as devices quickly move from one network to another, and it is desirable to minimize transition delays. For general purpose devices, RFC 4429 remains optional at this time.¶
[RFC7527] discusses enhanced DAD and describes an algorithm to automate the detection of looped-back IPv6 ND messages used by DAD. Nodes SHOULD implement this behavior where such detection is beneficial.¶
A node using Stateless Address Autoconfiguration [RFC4862] to form a globally unique IPv6 address that uses its MAC address to generate the IID will see that the IID remains the same on any visited network, even though the network prefix part changes. Thus, it is possible for a third-party device to track the activities of the node they communicate with, as that node moves around the network. Privacy Extensions for Stateless Address Autoconfiguration [RFC8981] address this concern by allowing nodes to configure an additional temporary address where the IID is effectively randomly generated. Privacy addresses are then used as source addresses for new communications initiated by the node.¶
General issues regarding privacy issues for IPv6 addressing are discussed in [RFC7721].¶
RFC 8981 SHOULD be supported. In some scenarios, such as dedicated servers in a data center, it provides limited or no benefit, or it may complicate network management. Thus, devices implementing this specification MUST provide a way for the end user to explicitly enable or disable the use of such temporary addresses.¶
Note that RFC 8981 can be used independently of traditional SLAAC or independently of SLAAC that is based on RFC 7217.¶
Implementers of RFC 8981 should be aware that certain addresses are reserved and should not be chosen for use as temporary addresses. Consult "Reserved IPv6 Interface Identifiers" [RFC5453] for more details.¶
DHCPv6 [RFC8415] can be used to obtain and configure addresses. In general, a network may provide for the configuration of addresses through SLAAC, DHCPv6, or both. There will be a wide range of IPv6 deployment models and differences in address assignment requirements, some of which may require DHCPv6 for stateful address assignment. Consequently, all hosts SHOULD implement address configuration via DHCPv6.¶
In the absence of observed Router Advertisement messages, IPv6 nodes MAY initiate DHCP to obtain IPv6 addresses and other configuration information, as described in Section 5.5.2 of [RFC4862].¶
Where devices are likely to be carried by users and attached to multiple visited networks, DHCPv6 client anonymity profiles SHOULD be supported as described in [RFC7844] to minimize the disclosure of identifying information. Section 5 of [RFC7844] describes operational considerations on the use of such anonymity profiles.¶
IPv6 nodes will invariably have multiple addresses configured simultaneously and thus will need to choose which addresses to use for which communications. The rules specified in the Default Address Selection for IPv6 document [RFC6724] MUST be implemented. [RFC8028] updates Rule 5.5 from [RFC6724]; implementations SHOULD implement this rule.¶
DNS is described in [RFC1034], [RFC1035], [RFC3363], and [RFC3596]. Not all nodes will need to resolve names; those that will never need to resolve DNS names do not need to implement resolver functionality. However, the ability to resolve names is a basic infrastructure capability on which applications rely, and most nodes will need to provide support. All nodes SHOULD implement stub-resolver [RFC1034] functionality, as in Section 5.3.1 of [RFC1034], with support for:¶
reverse addressing in ip6.arpa using PTR records [RFC3596]; and¶
Extension Mechanisms for DNS (EDNS(0)) [RFC6891] to allow for DNS packet sizes larger than 512 octets.¶
Those nodes are RECOMMENDED to support DNS security extensions [RFC4033] [RFC4034] [RFC4035].¶
A6 Resource Records [RFC2874] are classified as Historic per [RFC6563]. These were defined with Experimental status in [RFC3363].¶
DHCP [RFC8415] specifies a mechanism for IPv6 nodes to obtain address configuration information (see Section 6.5) and to obtain additional (non-address) configuration. If a host implementation supports applications or other protocols that require configuration that is only available via DHCP, hosts SHOULD implement DHCP. For specialized devices on which no such configuration need is present, DHCP may not be necessary.¶
An IPv6 node can use the subset of DHCP (described in [RFC8415]) to obtain other configuration information.¶
If an IPv6 node implements DHCP, it MUST implement the DNS options [RFC3646] as most deployments will expect that these options are available.¶
There is no defined DHCPv6 Gateway option.¶
Nodes using the Dynamic Host Configuration Protocol for IPv6 (DHCPv6) are thus expected to determine their default router information and on-link prefix information from received Router Advertisements.¶
Router Advertisement options have historically been limited to those that are critical to basic IPv6 functionality. Originally, DNS configuration was not included as an RA option, and DHCP was the recommended way to obtain DNS configuration information. Over time, the thinking surrounding such an option has evolved. It is now generally recognized that few nodes can function adequately without having access to a working DNS resolver; thus, a Standards Track document has been published to provide this capability [RFC8106].¶
Implementations MUST include support for the DNS RA option [RFC8106].¶
In IPv6, there are two main protocol mechanisms for propagating configuration information to hosts: RAs and DHCP. RA options have been restricted to those deemed essential for basic network functioning and for which all nodes are configured with exactly the same information. Examples include the Prefix Information Options, the MTU option, etc. On the other hand, DHCP has generally been preferred for configuration of more general parameters and for parameters that may be client specific. Generally speaking, however, there has been a desire to define only one mechanism for configuring a given option, rather than defining multiple (different) ways of configuring the same information.¶
One issue with having multiple ways to configure the same information is that interoperability suffers if a host chooses one mechanism but the network operator chooses a different mechanism. For "closed" environments, where the network operator has significant influence over what devices connect to the network and thus what configuration mechanisms they support, the operator may be able to ensure that a particular mechanism is supported by all connected hosts. In more open environments, however, where arbitrary devices may connect (e.g., a Wi-Fi hotspot), problems can arise. To maximize interoperability in such environments, hosts would need to implement multiple configuration mechanisms to ensure interoperability.¶
Multicast DNS (mDNS) and DNS-based Service Discovery (DNS-SD) are described in [RFC6762] and [RFC6763], respectively. These protocols, often collectively referred to as the 'Bonjour' protocols after their naming by Apple, provide the means for devices to discover services within a local link and, in the absence of a unicast DNS service, to exchange naming information.¶
Where devices are to be deployed in networks where service discovery would be beneficial, e.g., for users seeking to discover printers or display devices, mDNS and DNS-SD SHOULD be supported.¶
IPv6 nodes MAY support IPv4.¶
If an IPv6 node implements dual stack and tunneling, then [RFC4213] MUST be supported.¶
Software that allows users and operators to input IPv6 addresses in text form SHOULD support "A Recommendation for IPv6 Address Text Representation" [RFC5952].¶
There are a number of IPv6-related APIs. This document does not mandate the use of any, because the choice of API does not directly relate to on-the-wire behavior of protocols. Implementers, however, would be advised to consider providing a common API or reviewing existing APIs for the type of functionality they provide to applications.¶
"Basic Socket Interface Extensions for IPv6" [RFC3493] provides IPv6 functionality used by typical applications. Implementers should note that RFC 3493 has been picked up and further standardized by the Portable Operating System Interface (POSIX) [POSIX].¶
"Advanced Sockets Application Program Interface (API) for IPv6" [RFC3542] provides access to advanced IPv6 features needed by diagnostic and other more specialized applications.¶
"IPv6 Socket API for Source Address Selection" [RFC5014] provides facilities that allow an application to override the default Source Address Selection rules of [RFC6724].¶
"Socket Interface Extensions for Multicast Source Filters" [RFC3678] provides support for expressing source filters on multicast group memberships.¶
"Extension to Sockets API for Mobile IPv6" [RFC4584] provides application support for accessing and enabling Mobile IPv6 [RFC6275] features.¶
Mobile IPv6 [RFC6275] and associated specifications [RFC3776] [RFC4877] allow a node to change its point of attachment within the Internet, while maintaining (and using) a permanent address. All communication using the permanent address continues to proceed as expected even as the node moves around. The definition of Mobile IP includes requirements for the following types of nodes:¶
mobile nodes¶
correspondent nodes with support for route optimization¶
home agents¶
all IPv6 routers¶
At the present time, Mobile IP has seen only limited implementation and no significant deployment, partly because it originally assumed an IPv6-only environment rather than a mixed IPv4/IPv6 Internet. Additional work has been done to support mobility in mixed-mode IPv4 and IPv6 networks [RFC5555].¶
More usage and deployment experience is needed with mobility before any specific approach can be recommended for broad implementation in all hosts and routers. Consequently, Mobility Support in IPv6 [RFC6275], Mobile IPv6 Support for Dual Stack Hosts and Routers [RFC5555], and associated standards (such as Mobile IPv6 with IKEv2 and IPsec [RFC4877]) are considered a MAY at this time.¶
IPv6 for 3GPP [RFC7066] lists a snapshot of required IPv6 functionalities at the time the document was published that would need to be implemented, going above and beyond the recommendations in this document. Additionally, a 3GPP IPv6 Host MAY implement [RFC7278] to deliver IPv6 prefixes on the LAN link.¶
This section describes the security specification for IPv6 nodes.¶
Achieving security in practice is a complex undertaking. Operational procedures, protocols, key distribution mechanisms, certificate management approaches, etc., are all components that impact the level of security actually achieved in practice. More importantly, deficiencies or a poor fit in any one individual component can significantly reduce the overall effectiveness of a particular security approach.¶
IPsec can provide either end-to-end security between nodes or channel security (for example, via a site-to-site IPsec VPN), making it possible to provide secure communication for all (or a subset of) communication flows at the IP layer between pairs of Internet nodes. IPsec has two standard operating modes: Tunnel-mode and Transport-mode. In Tunnel-mode, IPsec provides network-layer security and protects an entire IP packet by encapsulating the original IP packet and then prepending a new IP header. In Transport-mode, IPsec provides security for the transport layer (and above) by encapsulating only the transport-layer (and above) portion of the IP packet (i.e., without adding a second IP header).¶
Although IPsec can be used with manual keying in some cases, such usage has limited applicability and is not recommended.¶
A range of security technologies and approaches proliferate today (e.g., IPsec, Transport Layer Security (TLS), Secure SHell (SSH), TLS VPNS, etc.). No single approach has emerged as an ideal technology for all needs and environments. Moreover, IPsec is not viewed as the ideal security technology in all cases and is unlikely to displace the others.¶
Previously, IPv6 mandated implementation of IPsec and recommended the key-management approach of IKE. RFC 6434 updated that recommendation by making support of the IPsec architecture [RFC4301] a SHOULD for all IPv6 nodes, and this document retains that recommendation. Note that the IPsec Architecture requires the implementation of both manual and automatic key management (e.g., Section 4.5 of [RFC4301]). Currently, the recommended automated key-management protocol to implement is IKEv2 [RFC7296].¶
This document recognizes that there exists a range of device types and environments where approaches to security other than IPsec can be justified. For example, special-purpose devices may support only a very limited number or type of applications, and an application-specific security approach may be sufficient for limited management or configuration capabilities. Alternatively, some devices may run on extremely constrained hardware (e.g., sensors) where the full IPsec Architecture is not justified.¶
Because most common platforms now support IPv6 and have it enabled by default, IPv6 security is an issue for networks that are ostensibly IPv4 only; see [RFC7123] for guidance on this area.¶
"Security Architecture for the Internet Protocol" [RFC4301] SHOULD be supported by all IPv6 nodes. Note that the IPsec Architecture requires the implementation of both manual and automatic key management (e.g., Section 4.5 of [RFC4301]). Currently, the default automated key-management protocol to implement is IKEv2. As required in [RFC4301], IPv6 nodes implementing the IPsec Architecture MUST implement ESP [RFC4303] and MAY implement AH [RFC4302].¶
The current set of mandatory-to-implement algorithms for the IPsec Architecture are defined in Cryptographic Algorithm Implementation Requirements for ESP and AH [RFC8221]. IPv6 nodes implementing the IPsec Architecture MUST conform to the requirements in [RFC8221]. Preferred cryptographic algorithms often change more frequently than security protocols. Therefore, implementations MUST allow for migration to new algorithms, as RFC 8221 is replaced or updated in the future.¶
The current set of mandatory-to-implement algorithms for IKEv2 are defined in Cryptographic Algorithm Implementation Requirements for ESP and AH [RFC8247]. IPv6 nodes implementing IKEv2 MUST conform to the requirements in [RFC8247] and/or any future updates or replacements to [RFC8247].¶
This section defines general host considerations for IPv6 nodes that act as routers. Currently, this section does not discuss detailed routing-specific requirements. For the case of typical home routers, [RFC7084] defines basic requirements for customer edge routers.¶
The IPv6 Router Alert option [RFC2711] is an optional IPv6 Hop-by-Hop Header that is used in conjunction with some protocols (e.g., RSVP [RFC2205] or Multicast Listener Discovery (MLDv2) [RFC3810]). The Router Alert option will need to be implemented whenever such protocols that mandate its use are implemented. See Section 5.11.¶
Sending Router Advertisements and processing Router Solicitations MUST be supported.¶
Routers SHOULD support [RFC9131] to avoid packet lost.¶
Section 7 of [RFC6275] includes some mobility-specific extensions to Neighbor Discovery. Routers SHOULD implement Sections 7.3 and 7.5, even if they do not implement home agent functionality.¶
A single DHCP server ([RFC8415] or [RFC4862]) can provide configuration information to devices directly attached to a shared link, as well as to devices located elsewhere within a site. Communication between a client and a DHCP server located on different links requires the use of DHCP relay agents on routers.¶
In simple deployments, consisting of a single router and either a single LAN or multiple LANs attached to the single router, together with a WAN connection, a DHCP server embedded within the router is one common deployment scenario (e.g., [RFC7084]). There is no need for relay agents in such scenarios.¶
In more complex deployment scenarios, such as within enterprise or service provider networks, the use of DHCP requires some level of configuration, in order to configure relay agents, DHCP servers, etc. In such environments, the DHCP server might even be run on a traditional server, rather than as part of a router.¶
Because of the wide range of deployment scenarios, support for DHCP server functionality on routers is optional. However, routers targeted for deployment within more complex scenarios (as described above) SHOULD support relay agent functionality. Note that "Basic Requirements for IPv6 Customer Edge Routers" [RFC7084] requires implementation of a DHCPv6 server function in IPv6 Customer Edge (CE) routers.¶
Forwarding nodes MUST conform to BCP 198 [RFC7608]; thus, IPv6 implementations of nodes that may forward packets MUST conform to the rules specified in Section 5.1 of [RFC4632].¶
The focus of this document is general IPv6 nodes. In this section, we briefly discuss considerations for constrained devices.¶
In the case of constrained nodes, with limited CPU, memory, bandwidth or power, support for certain IPv6 functionality may need to be considered due to those limitations. While the requirements of this document are RECOMMENDED for all nodes, including constrained nodes, compromises may need to be made in certain cases. Where such compromises are made, the interoperability of devices should be strongly considered, particularly where this may impact other nodes on the same link, e.g., only supporting MLDv1 will affect other nodes.¶
The IETF 6LowPAN (IPv6 over Low-Power Wireless Personal Area Network) WG produced six RFCs, including a general overview and problem statement [RFC4919] (the means by which IPv6 packets are transmitted over IEEE 802.15.4 networks [RFC4944] and ND optimizations for that medium [RFC6775]).¶
IPv6 nodes that are battery powered SHOULD implement the recommendations in [RFC7772].¶
Network management MAY be supported by IPv6 nodes. However, for IPv6 nodes that are embedded devices, network management may be the only possible way of controlling these nodes.¶
Existing network management protocols include SNMP [RFC3411], NETCONF [RFC6241], and RESTCONF [RFC8040].¶
The obsoleted status of various IPv6-specific MIB modules is clarified in [RFC8096].¶
The following two MIB modules SHOULD be supported by nodes that support an SNMP agent.¶
The IP Forwarding Table MIB [RFC4292] SHOULD be supported by nodes that support an SNMP agent.¶
The IP MIB [RFC4293] SHOULD be supported by nodes that support an SNMP agent.¶
The Interface MIB [RFC2863] SHOULD be supported by nodes that support an SNMP agent.¶
The following YANG data models SHOULD be supported by nodes that support a NETCONF or RESTCONF agent.¶
The IP Management YANG Model [RFC8344] SHOULD be supported by nodes that support NETCONF or RESTCONF.¶
The Interface Management YANG Model [RFC8343] SHOULD be supported by nodes that support NETCONF or RESTCONF.¶
This document does not directly affect the security of the Internet, beyond the security considerations associated with the individual protocols.¶
Security is also discussed in Section 13 above.¶
This document has no IANA actions.¶
This section highlights the changes since RFC 8504. It is not a complete list of all changes.¶
Updated obsoleted RFCs including 3315 and 3736 (both to 8415) and 4941 to 8981. RFC 793 has been obsoleted by 9293 but the latter does not include the the robustness principle for which RFC 793 is cited in this document.¶
Added support for RFC 9131.¶
Type C host from RFC 4191 is went from a SHOULD to MUST.¶
There have been many editorial clarifications as well as significant additions and updates. While this section highlights some of the changes, readers should not rely on this section for a comprehensive list of all changes.¶
Restructured sections.¶
Added 6LoWPAN to link layers as it has some deployment.¶
Removed the Downstream-on-Demand (DoD) IPv6 Profile as it hasn't been updated.¶
Updated MLDv2 support to a MUST since nodes are restricted if MLDv1 is used.¶
Required DNS RA options so SLAAC-only devices can get DNS; RFC 8106 is a MUST.¶
Required RFC 3646 DNS Options for DHCPv6 implementations.¶
Added RESTCONF and NETCONF as possible options to network management.¶
Added a section on constrained devices.¶
Added text on RFC 7934 to address availability to hosts (SHOULD).¶
Added text on RFC 7844 for anonymity profiles for DHCPv6 clients.¶
Added mDNS and DNS-SD as updated service discovery.¶
Added RFC 8028 as a SHOULD as a method for solving a multi-prefix network.¶
Added ECN RFC 3168 as a SHOULD.¶
Added reference to RFC 7123 for security over IPv4-only networks.¶
Removed Jumbograms (RFC 2675) as they aren't deployed.¶
Updated obsoleted RFCs to the new version of the RFC, including RFCs 2460, 1981, 7321, and 4307.¶
Added RFC 7772 for power consumptions considerations.¶
Added why /64 boundaries for more detail --- RFC 7421.¶
Added a unique IPv6 prefix per host to support currently deployed IPv6 networks.¶
Clarified RFC 7066 was a snapshot for 3GPP.¶
Updated RFC 4191 as a MUST and the Type C Host as a SHOULD as they help solve multi-prefix problems.¶
Removed IPv6 over ATM since there aren't many deployments.¶
Added a note in Section 6.6 for Rule 5.5 from RFC 6724.¶
Added MUST for BCP 198 for forwarding IPv6 packets.¶
Added a reference to RFC 8064 for stable address creation.¶
Added text on the protection from excessive extension header options.¶
Added text on the dangers of 1280 MTU UDP, especially with regard to DNS traffic.¶
Added text to clarify RFC 8200 behavior for unrecognized extension headers or unrecognized ULPs.¶
Removed dated email addresses from design team acknowledgements for [RFC4294].¶
There have been many editorial clarifications as well as significant additions and updates. While this section highlights some of the changes, readers should not rely on this section for a comprehensive list of all changes.¶
Updated the Introduction to indicate that this document is an applicability statement and is aimed at general nodes.¶
Significantly updated the section on mobility protocols; added references and downgraded previous SHOULDs to MAYs.¶
Changed the Sub-IP Layer section to just list relevant RFCs, and added some more RFCs.¶
Added a section on SEND (it is a MAY).¶
Revised the section on Privacy Extensions to add more nuance to the recommendation.¶
Completely revised the IPsec/IKEv2 section, downgrading the overall recommendation to a SHOULD.¶
Upgraded recommendation of DHCPv6 to a SHOULD.¶
Added a background section on DHCP versus RA options, added a SHOULD recommendation for DNS configuration via RAs (RFC 6106), and cleaned up the DHCP recommendations.¶
Added the recommendation that routers implement Sections 7.3 and 7.5 of [RFC6275].¶
Added a pointer to subnet clarification document [RFC5942].¶
Added text that "IPv6 Host-to-Router Load Sharing" [RFC4311] SHOULD be implemented.¶
Added reference to [RFC5722] (Overlapping Fragments), and made it a MUST to implement.¶
Made "A Recommendation for IPv6 Address Text Representation" [RFC5952] a SHOULD.¶
Removed the mention of delegation name (DNAME) from the discussion about [RFC3363].¶
Numerous updates to reflect newer versions of IPv6 documents, including [RFC3596], [RFC4213], [RFC4291], and [RFC4443].¶
Removed discussion of "Managed" and "Other" flags in RAs. There is no consensus at present on how to process these flags, and discussion of their semantics was removed in the most recent update of Stateless Address Autoconfiguration [RFC4862].¶
Added many more references to optional IPv6 documents.¶
Made "A Recommendation for IPv6 Address Text Representation" [RFC5952] a SHOULD.¶
Updated the MLD section to include reference to Lightweight MLD [RFC5790].¶
Added a SHOULD recommendation for "Default Router Preferences and More-Specific Routes" [RFC4191].¶
Acknowledgments (Current Document)¶
The authors would like to thank Brian Carpenter, Dave Thaler, Tom Herbert, Erik Kline, Mohamed Boucadair, and Michayla Newcombe for their contributions and many members of the 6man WG for the inputs they gave.¶
Authors and Acknowledgments from RFC 6434¶
RFC 6434 was authored by Ed Jankiewicz, John Loughney, and Thomas Narten.¶
The authors of RFC 6434 thank Hitoshi Asaeda, Brian Carpenter, Tim Chown, Ralph Droms, Sheila Frankel, Sam Hartman, Bob Hinden, Paul Hoffman, Pekka Savola, Yaron Sheffer, and Dave Thaler for their comments. In addition, the authors thank Mark Andrews for comments and corrections on DNS text and Alfred Hoenes for tracking the updates to various RFCs.¶
Authors and Acknowledgments from RFC 4294¶
RFC 4294 was written by the IPv6 Node Requirements design team, which had the following members: Jari Arkko, Marc Blanchet, Samita Chakrabarti, Alain Durand, Gerard Gastaud, Jun-ichiro Itojun Hagino, Atsushi Inoue, Masahiro Ishiyama, John Loughney, Rajiv Raghunarayan, Shoichi Sakane, Dave Thaler, and Juha Wiljakka.¶
The authors of RFC 4294 thank Ran Atkinson, Jim Bound, Brian Carpenter, Ralph Droms, Christian Huitema, Adam Machalek, Thomas Narten, Juha Ollila, and Pekka Savola for their comments.¶