RFC 2373

IP Version 6 Addressing Architecture

📅 published July 1998, from ietf

1.0 Introduction 🌸

Okay bestie, this is RFC 2373 🎀, the IPv6 addressing architecture. It's a Standards Track doc from the Network Working Group, published in July 1998 ✨. R. Hinden (Nokia) and S. Deering (Cisco Systems) wrote it 💖, and it obsoletes RFC 1884 🫶, so 1884 is out and this is the new tea.

Request for Comments: 2373

Category: Standards Track

Authors: R. Hinden (Nokia), S. Deering (Cisco Systems)

Date: July 1998

Obsoletes: RFC 1884

Status of this Memo ✨

This doc is an Internet standards track protocol for the Internet community, and it's asking for discussion and suggestions for improvements 💅. For the standardization state and status of this protocol, check the current edition of the "Internet Official Protocol Standards" (STD 1). Distribution of this memo is unlimited 🦋.

Copyright (C) The Internet Society (1998). All Rights Reserved.

Abstract 🧚

This spec defines the addressing architecture of the IP Version 6 protocol [IPV6]. It covers the IPv6 addressing model, text representations of IPv6 addresses, and definitions of IPv6 unicast, anycast, and multicast addresses. It also covers the addresses an IPv6 node is required to have 🔥.

Table of Contents 🍓

1. Introduction.................................................2
2. IPv6 Addressing..............................................2
   2.1 Addressing Model.........................................3
   2.2 Text Representation of Addresses.........................3
   2.3 Text Representation of Address Prefixes..................5
   2.4 Address Type Representation..............................6
   2.5 Unicast Addresses........................................7
     2.5.1 Interface Identifiers................................8
     2.5.2 The Unspecified Address..............................9
     2.5.3 The Loopback Address.................................9
     2.5.4 IPv6 Addresses with Embedded IPv4 Addresses.........10
     2.5.5 NSAP Addresses......................................10
     2.5.6 IPX Addresses.......................................10
     2.5.7 Aggregatable Global Unicast Addresses...............11
     2.5.8 Local-use IPv6 Unicast Addresses....................11
   2.6 Anycast Addresses.......................................12
     2.6.1 Required Anycast Address............................13
   2.7 Multicast Addresses.....................................14

     2.7.1 Pre-Defined Multicast Addresses.....................15
     2.7.2 Assignment of New IPv6 Multicast Addresses..........17
   2.8 A Node's Required Addresses.............................17
3. Security Considerations.....................................18
APPENDIX A: Creating EUI-64 based Interface Identifiers........19
APPENDIX B: ABNF Description of Text Representations...........22
APPENDIX C: CHANGES FROM RFC-1884..............................23
REFERENCES.....................................................24
AUTHORS' ADDRESSES.............................................25
FULL COPYRIGHT STATEMENT.......................................26

1.0 Introduction 🌸

This spec defines the addressing architecture of the IP Version 6 protocol. It includes a detailed description of the address formats currently defined for IPv6 [IPV6] 💖.

The authors want to thank Paul Francis, Scott Bradner, Jim Bound, Brian Carpenter, Matt Crawford, Deborah Estrin, Roger Fajman, Bob Fink, Peter Ford, Bob Gilligan, Dimitry Haskin, Tom Harsch, Christian Huitema, Tony Li, Greg Minshall, Thomas Narten, Erik Nordmark, Yakov Rekhter, Bill Simpson, and Sue Thomson 🫶.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC 2119]. These are the main-character words of the whole doc 👑, so read them carefully.

2.0 IPv6 Addressing 💅

IPv6 addresses are 128-bit identifiers for interfaces and sets of interfaces. There are three types of addresses 🦋:

  Unicast:   An identifier for a single interface.  A packet sent to
             a unicast address is delivered to the interface
             identified by that address.

  Anycast:   An identifier for a set of interfaces (typically
             belonging to different nodes).  A packet sent to an
             anycast address is delivered to one of the interfaces
             identified by that address (the "nearest" one, according
             to the routing protocols' measure of distance).

  Multicast: An identifier for a set of interfaces (typically
             belonging to different nodes).  A packet sent to a
             multicast address is delivered to all interfaces
             identified by that address.

There are no broadcast addresses in IPv6. Multicast addresses took over that job 🌈.

In this document, fields in addresses get a specific name, like "subscriber". When that name is followed by the word "ID" (as in "subscriber ID"), it means the contents of the named field. When it's followed by the word "prefix" (as in "subscriber prefix"), it means the whole address up to and including that field 🍓.

In IPv6, all zeros and all ones are legal values for any field, unless something specifically excludes them. Prefixes can contain zero-valued fields or end in zeros, so don't panic when you see them 🙈.

2.1 Addressing Model 🦋

IPv6 addresses of every type get assigned to interfaces, not nodes 💖. A unicast address points at one single interface. Since each interface belongs to exactly one node, any of that node's interfaces' unicast addresses can be used to identify the node ✨.

Every interface is required to have at least one link-local unicast address (see section 2.8 for the additional required addresses) 💅. An interface can also be given several IPv6 addresses of any type (unicast, anycast, and multicast) or scope. Unicast addresses with a scope bigger than link-scope aren't needed for interfaces that never serve as the origin or destination of any IPv6 packets going to or coming from non-neighbors. This is sometimes handy for point-to-point interfaces 🫶. There is one exception to this addressing model:

  A unicast address or a set of unicast addresses may be assigned to
  multiple physical interfaces if the implementation treats the
  multiple physical interfaces as one interface when presenting it to
  the internet layer.  This is useful for load-sharing over multiple
  physical interfaces.

Right now IPv6 keeps the IPv4 model where a subnet prefix is tied to one link. Multiple subnet prefixes can be assigned to the same link 🌸.

2.2 Text Representation of Addresses 💜

There are three conventional forms for writing IPv6 addresses as text strings 🧚:

  1. The preferred form is x:x:x:x:x:x:x:x, where the 'x's are the hexadecimal values of the eight 16-bit pieces of the address. Examples:
   FEDC:BA98:7654:3210:FEDC:BA98:7654:3210

   1080:0:0:0:8:800:200C:417A

You don't have to write the leading zeros in an individual field, but there must be at least one numeral in every field (except for the case described in 2.) 🔥.

  1. Lots of IPv6 addresses end up with long strings of zero bits, because of how certain styles of addresses get allocated. To make writing those easier, there's a special syntax that compresses the zeros. Using "::" stands for multiple groups of 16 bits of zeros 🍓. The "::" can only show up once in an address. It can also be used to compress leading and/or trailing zeros in an address ✨.

For example, these addresses:

   1080:0:0:0:8:800:200C:417A  a unicast address
   FF01:0:0:0:0:0:0:101        a multicast address
   0:0:0:0:0:0:0:1             the loopback address
   0:0:0:0:0:0:0:0             the unspecified addresses

can be written as:

   1080::8:800:200C:417A       a unicast address
   FF01::101                   a multicast address
   ::1                         the loopback address
   ::                          the unspecified addresses
  1. There's also an alternative form that's sometimes more convenient in a mixed environment of IPv4 and IPv6 nodes: x:x:x:x:x:x:d.d.d.d. The 'x's are the hexadecimal values of the six high-order 16-bit pieces of the address, and the 'd's are the decimal values of the four low-order 8-bit pieces (the standard IPv4 representation) 🦋. Examples:
   0:0:0:0:0:0:13.1.68.3

   0:0:0:0:0:FFFF:129.144.52.38

or in compressed form:

   ::13.1.68.3

   ::FFFF:129.144.52.38

2.3 Text Representation of Address Prefixes 🎀

Writing IPv6 address prefixes looks a lot like how IPv4 prefixes get written in CIDR notation 💖. An IPv6 address prefix is written like this:

   ipv6-address/prefix-length

where

   ipv6-address    is an IPv6 address in any of the notations listed
                   in section 2.2.

   prefix-length   is a decimal value specifying how many of the
                   leftmost contiguous bits of the address comprise
                   the prefix.

For example, these are all legal ways to write the 60-bit prefix 12AB00000000CD3 (hexadecimal) ✨:

   12AB:0000:0000:CD30:0000:0000:0000:0000/60
   12AB::CD30:0:0:0:0/60
   12AB:0:0:CD30::/60

These are not legal ways to write the prefix above 🙈:

   12AB:0:0:CD3/60   may drop leading zeros, but not trailing zeros,
                     within any 16-bit chunk of the address

   12AB::CD30/60     address to left of "/" expands to
                     12AB:0000:0000:0000:0000:000:0000:CD30

   12AB::CD3/60      address to left of "/" expands to
                     12AB:0000:0000:0000:0000:000:0000:0CD3

When you write both a node address and a prefix of that node address (like the node's subnet prefix), you can combine them like this 💅:

   the node address      12AB:0:0:CD30:123:4567:89AB:CDEF
   and its subnet number 12AB:0:0:CD30::/60

   can be abbreviated as 12AB:0:0:CD30:123:4567:89AB:CDEF/60

2.4 Address Type Representation ✨

okay so the type of an IPv6 address is basically written in its first few bits 🔍 That leading chunk is called the Format Prefix (FP), and it's variable length, so it can be different sizes depending on the type. Here's the initial allocation of those prefixes 👇

Allocation                            Prefix         Fraction of
                                      (binary)       Address Space
-----------------------------------   --------       -------------
Reserved                              0000 0000      1/256
Unassigned                            0000 0001      1/256

Reserved for NSAP Allocation          0000 001       1/128
Reserved for IPX Allocation           0000 010       1/128

Unassigned                            0000 011       1/128
Unassigned                            0000 1         1/32
Unassigned                            0001           1/16

Aggregatable Global Unicast Addresses 001            1/8
Unassigned                            010            1/8
Unassigned                            011            1/8
Unassigned                            100            1/8
Unassigned                            101            1/8
Unassigned                            110            1/8

Unassigned                            1110           1/16
Unassigned                            1111 0         1/32
Unassigned                            1111 10        1/64
Unassigned                            1111 110       1/128
Unassigned                            1111 1110 0    1/512

Link-Local Unicast Addresses          1111 1110 10   1/1024
Site-Local Unicast Addresses          1111 1110 11   1/1024

Multicast Addresses                   1111 1111      1/256

Notes: 🎀

  1. The "unspecified address" (see section 2.5.2), the loopback address (see section 2.5.3), and the IPv6 Addresses with Embedded IPv4 Addresses (see section 2.5.4), are assigned out of the 0000 0000 format prefix space.
  2. The format prefixes 001 through 111, except for Multicast Addresses (1111 1111), are all required to have 64-bit interface identifiers in EUI-64 format. See section 2.5.1 for definitions.

This allocation supports the direct allocation of aggregation addresses, local use addresses, and multicast addresses 💖 Space is reserved for NSAP addresses and IPX addresses, and the rest of the address space is unassigned for future use. That leftover space can be used to expand existing stuff (like extra aggregatable addresses, etc.) or for brand new uses (like separate locators and identifiers) 🧚 Only fifteen percent of the address space is initially allocated, so the remaining 85% is reserved for future use. Iconic planning, honestly 👑

Now, how do you tell unicast from multicast? Easy 💅 Look at the high-order octet. A value of FF (11111111) means it's a multicast address, and any other value means it's a unicast address. Anycast addresses are taken from the unicast address space, so they don't look any different from unicast addresses syntactically. Sneaky but make it make sense 🙈

2.5 Unicast Addresses 🦋

IPv6 unicast addresses are aggregatable with contiguous bit-wise masks, kind of like IPv4 addresses under Class-less Interdomain Routing [CIDR]. The vibes are very familiar ✨

There are several forms of unicast address assignment in IPv6. That includes the global aggregatable global unicast address, the NSAP address, the IPX hierarchical address, the site-local address, the link-local address, and the IPv4-capable host address. More address types can be defined later on, so this list is not the end of the story 🌈

How much a node knows about what's inside an IPv6 address really depends on what job the node has, like whether it's a host or a router 🎀 At the very least, a node can treat unicast addresses (including its own) as having no internal structure at all, just one 128-bit blob:

|                           128 bits                              |
+-----------------------------------------------------------------+
|                          node address                           |
+-----------------------------------------------------------------+

A slightly more sophisticated host (still pretty simple, though) might also know about the subnet prefix(es) for the link(s) it's attached to. Different addresses can have different values for n here:

|                         n bits                 |   128-n bits   |
+------------------------------------------------+----------------+
|                   subnet prefix                | interface ID   |
+------------------------------------------------+----------------+

Hosts that are even more sophisticated may know about other hierarchical boundaries inside the unicast address 🧠 A very simple router might not know anything about the internal structure of IPv6 unicast addresses, but routers will more generally know about one or more of the hierarchical boundaries so they can run their routing protocols. Which boundaries a router knows will be different from router to router, because it depends on what position that router holds in the routing hierarchy 🔥

2.5.1 Interface Identifiers 🫶

Interface identifiers in IPv6 unicast addresses are used to identify interfaces on a link. They're required to be unique on that link, and they may also be unique over a broader scope. In lots of cases, an interface's identifier is the same as that interface's link-layer address. The same interface identifier can be used on multiple interfaces of a single node too 🍓

Heads up: using the same interface identifier on multiple interfaces of one node does not affect that identifier's global uniqueness, or the global uniqueness of each IPv6 address created with that identifier. Iykyk 🙈

In a number of the format prefixes (see section 2.4), Interface IDs are required to be 64 bits long and built in IEEE EUI-64 format [EUI64]. EUI-64 based Interface identifiers can have global scope when a global token is available (for example, an IEEE 48-bit MAC), or local scope when a global token isn't available (for example, serial links, tunnel end-points, etc.) 💜 It's required that the "u" bit (universal/local bit in IEEE EUI-64 terminology) be inverted when forming the interface identifier from the EUI-64. The "u" bit is set to one (1) to indicate global scope, and set to zero (0) to indicate local scope. The first three octets in binary of an EUI-64 identifier look like this:

    0       0 0       1 1       2
   |0       7 8       5 6       3|
   +----+----+----+----+----+----+
   |cccc|ccug|cccc|cccc|cccc|cccc|
   +----+----+----+----+----+----+

written in Internet standard bit-order, where "u" is the universal/local bit, "g" is the individual/group bit, and "c" are the bits of the company_id. Appendix A: "Creating EUI-64 based Interface Identifiers" has examples of how to create different EUI-64 based interface identifiers 📚

So why flip the "u" bit? 🤔 The whole point is to make life easier for system administrators who need to hand-configure local scope identifiers when hardware tokens aren't available. That's expected to be the case for serial links, tunnel end-points, etc. The alternative would have been identifiers like 0200:0:0:1, 0200:0:0:2, and so on, instead of the much simpler ::1, ::2, etc. Way less cute 💀

Meanwhile, the universal/local bit in the IEEE EUI-64 identifier exists so future technology can take advantage of interface identifiers with global scope 🚀

The details of forming interface identifiers are defined in the appropriate "IPv6 over <link>" specification, like "IPv6 over Ethernet" [ETHER], "IPv6 over FDDI" [FDDI], and so on. Go check those out, bestie 🎀

2.5.2 The Unspecified Address 🌸

The address 0:0:0:0:0:0:0:0 is called the unspecified address. It must never be assigned to any node. It means "no address at all" 🫥 One example of when you'd use it: in the Source Address field of any IPv6 packets sent by an initializing host before it has learned its own address.

The unspecified address must not be used as the destination address of IPv6 packets, or in IPv6 Routing Headers. Absolutely not, zero chance 🙅‍♀️

2.5.3 The Loopback Address 🔁

The unicast address 0:0:0:0:0:0:0:1 is called the loopback address. A node may use it to send an IPv6 packet to itself. It may never be assigned to any physical interface, but you can think of it as being associated with a virtual interface (for example, the loopback interface) 💻

The loopback address must not be used as the source address in IPv6 packets that are sent outside of a single node. And an IPv6 packet with a destination address of loopback must never be sent outside of a single node, and must never be forwarded by an IPv6 router. Keep it at home, loopback 🏠✨

2.5.4 IPv6 Addresses with Embedded IPv4 Addresses 🌈

The IPv6 transition mechanisms [TRAN] include a technique for hosts and routers to dynamically tunnel IPv6 packets over IPv4 routing infrastructure 🚇 Nodes that use this technique are assigned special IPv6 unicast addresses that carry an IPv4 address in the low-order 32 bits. This type of address is called an "IPv4-compatible IPv6 address", and it looks like this:

|                80 bits               | 16 |      32 bits        |
+--------------------------------------+--------------------------+
|0000..............................0000|0000|    IPv4 address     |
+--------------------------------------+----+---------------------+

There's a second type of IPv6 address that holds an embedded IPv4 address too 🦋 It's used to represent the addresses of IPv4-only nodes (those that do not support IPv6) as IPv6 addresses. This type is called an "IPv4-mapped IPv6 address", and it's formatted like this:

|                80 bits               | 16 |      32 bits        |
+--------------------------------------+--------------------------+
|0000..............................0000|FFFF|    IPv4 address     |
+--------------------------------------+----+---------------------+

2.5.5 NSAP Addresses 🌷

The mapping of NSAP addresses into IPv6 addresses is defined in [NSAP]. Here's the tea: this document recommends that network implementors who have planned or deployed an OSI NSAP addressing plan, and who want to deploy or transition to IPv6, should redesign a native IPv6 addressing plan to meet their needs 🧁 But it also defines a set of mechanisms for supporting OSI NSAP addressing in an IPv6 network. If that support is required, these are the mechanisms that must be used. This document also defines a mapping of IPv6 addresses within the OSI address format, in case that's needed too 💅

2.5.6 IPX Addresses 🍒

The mapping of IPX addresses into IPv6 addresses goes like this:

|   7   |                   121 bits                              |
+-------+---------------------------------------------------------+
|0000010|                 to be defined                           |
+-------+---------------------------------------------------------+

The draft definition, motivation, and usage are under study. 🧚

2.5.7. Aggregatable Global Unicast Addresses ✨

okay so the global aggregatable global unicast address is defined in [AGGR] 💖 This format is built to support BOTH the current provider based aggregation AND a brand new type of aggregation called exchanges 🦋 The combo means efficient routing aggregation for sites that connect straight to providers and for sites that connect to exchanges. Sites get to pick which aggregation point they hook up to, which is so iconic of them 👑

The IPv6 aggregatable global unicast address format looks like this:

   | 3|  13 | 8 |   24   |   16   |          64 bits               |
   +--+-----+---+--------+--------+--------------------------------+
   |FP| TLA |RES|  NLA   |  SLA   |         Interface ID           |
   |  | ID  |   |  ID    |  ID    |                                |
   +--+-----+---+--------+--------+--------------------------------+

Here's what each piece means 🎀

  • 001 is the Format Prefix (3 bit) for Aggregatable Global Unicast Addresses
  • TLA ID is the Top-Level Aggregation Identifier
  • RES is reserved for future use
  • NLA ID is the Next-Level Aggregation Identifier
  • SLA ID is the Site-Level Aggregation Identifier
  • INTERFACE ID is the Interface Identifier

The contents, field sizes, and assignment rules are defined in [AGGR]. Go read that one if you want the full details, bestie 📚🫶

2.5.8. Local-Use IPv6 Unicast Addresses 🌸

There are two types of local-use unicast addresses defined, and they're called Link-Local and Site-Local 🧚 Link-Local is for use on a single link, and Site-Local is for use inside a single site. Here's the Link-Local format:

   |   10     |
   |  bits    |        54 bits          |          64 bits           |
   +----------+-------------------------+----------------------------+
   |1111111010|           0             |       interface ID         |
   +----------+-------------------------+----------------------------+

Link-Local addresses are designed to be used for addressing on a single link for purposes such as auto-address configuration, neighbor discovery, or when no routers are present. Main character energy 💅

Routers must not forward any packets with link-local source or destination addresses to other links. Not me forwarding link-local stuff off the link 🙈

Now the Site-Local format:

   |   10     |
   |  bits    |   38 bits   |  16 bits  |         64 bits            |
   +----------+-------------+-----------+----------------------------+
   |1111111011|    0        | subnet ID |       interface ID         |
   +----------+-------------+-----------+----------------------------+

Site-Local addresses are designed to be used for addressing inside of a site without the need for a global prefix. No prefix drama needed 💜

Routers must not forward any packets with site-local source or destination addresses outside of the site. Periodt 🔥

2.6. Anycast Addresses 🦋

An IPv6 anycast address is an address that gets assigned to more than one interface, usually on different nodes. The fun property is that a packet sent to an anycast address gets routed to the "nearest" interface that has that address, based on the routing protocols' measure of distance. It's like the closest friend who always shows up first 🍓

Anycast addresses are allocated from the unicast address space, using any of the defined unicast address formats. That means anycast addresses are syntactically indistinguishable from unicast addresses, so they look exactly the same ngl 🙃 When a unicast address gets assigned to more than one interface and turns into an anycast address, the nodes it's assigned to must be explicitly configured to know that it is an anycast address. They have to be told, no vibes-only detection 🫠

For any assigned anycast address, there is a longest address prefix P that identifies the topological region where all the interfaces belonging to that anycast address live. Inside the region identified by P, each member of the anycast set must be advertised as a separate entry in the routing system (commonly called a "host route"). Outside the region identified by P, the anycast address may be aggregated into the routing advertisement for prefix P. Aggregation is the whole goal, bestie ✨

Note that, in the worst case, the prefix P of an anycast set may be the null prefix. In other words, the members of the set may have no topological locality at all 😭 In that case, the anycast address must be advertised as a separate routing entry throughout the entire internet, and that gives a severe scaling limit on how many such "global" anycast sets can be supported. So it's expected that support for global anycast sets may be unavailable or very restricted. Major limits, I know 💀

One expected use of anycast addresses is to identify the set of routers belonging to an organization that provides internet service. These addresses could be used as intermediate addresses in an IPv6 Routing header, to cause a packet to be delivered via a particular aggregation or sequence of aggregations. Some other possible uses are identifying the set of routers attached to a particular subnet, or the set of routers providing entry into a particular routing domain 🌈

There is little experience with widespread, arbitrary use of internet anycast addresses, and there are some known complications and hazards when using them in their full generality [ANYCST]. Until more experience has been gained and solutions have been agreed upon for those problems, the following restrictions are imposed on IPv6 anycast addresses 🔒

  • An anycast address must not be used as the source address of an IPv6 packet.
  • An anycast address must not be assigned to an IPv6 host. That is, it may be assigned to an IPv6 router only. Routers only, babes 🧚

2.6.1. Required Anycast Address 💖

The Subnet-Router anycast address is predefined. Its format is as follows:

   |                         n bits                 |   128-n bits   |
   +------------------------------------------------+----------------+
   |                   subnet prefix                | 00000000000000 |
   +------------------------------------------------+----------------+

The "subnet prefix" in an anycast address is the prefix that identifies a specific link. This anycast address is syntactically the same as a unicast address for an interface on the link, with the interface identifier set to zero. Same look, different job 🎀

Packets sent to the Subnet-Router anycast address will be delivered to one router on the subnet. All routers are required to support the Subnet-Router anycast addresses for the subnets they have interfaces on. Every router has to show up for this one 🫶

The subnet-router anycast address is intended to be used for applications where a node needs to communicate with one of a set of routers on a remote subnet. For example, this happens when a mobile host needs to communicate with one of the mobile agents on its "home" subnet 🏠

2.7. Multicast Addresses 🌈

An IPv6 multicast address is an identifier for a group of nodes. A node may belong to any number of multicast groups. Multicast addresses have the following format:

   |   8    |  4 |  4 |                  112 bits                   |
   +------ -+----+----+---------------------------------------------+
   |11111111|flgs|scop|                  group ID                   |
   +--------+----+----+---------------------------------------------+

11111111 at the start of the address identifies it as a multicast address. That's the giveaway 👀

The flgs field is a set of 4 flags, laid out like this:

                                    +-+-+-+-+
      flgs is a set of 4 flags:     |0|0|0|T|
                                    +-+-+-+-+
  • The high-order 3 flags are reserved, and must be initialized to 0.
  • T = 0 indicates a permanently-assigned ("well-known") multicast address, assigned by the global internet numbering authority. Forever vibes ⭐
  • T = 1 indicates a non-permanently-assigned ("transient") multicast address. Temporary era 🦋

The scop field is a 4-bit multicast scope value used to limit the scope of the multicast group. The values are:

      0  reserved
      1  node-local scope
      2  link-local scope
      3  (unassigned)
      4  (unassigned)
      5  site-local scope
      6  (unassigned)
      7  (unassigned)
      8  organization-local scope
      9  (unassigned)
      A  (unassigned)
      B  (unassigned)
      C  (unassigned)
      D  (unassigned)
      E  global scope
      F  reserved

The group ID identifies the multicast group, either permanent or transient, within the given scope 💜

The "meaning" of a permanently-assigned multicast address is independent of the scope value. For example, say the "NTP servers group" is assigned a permanent multicast address with a group ID of 101 (hex). Then:

  • FF01:0:0:0:0:0:0:101 means all NTP servers on the same node as the sender. Same device, bestie 🙈
  • FF02:0:0:0:0:0:0:101 means all NTP servers on the same link as the sender. Same neighborhood 🏘️
  • FF05:0:0:0:0:0:0:101 means all NTP servers at the same site as the sender. Same campus 🎓
  • FF0E:0:0:0:0:0:0:101 means all NTP servers in the internet. Everyone, everywhere 🌍✨

Non-permanently-assigned multicast addresses are only meaningful within a given scope. For example, a group identified by the non-permanent, site-local multicast address FF15:0:0:0:0:0:0:101 at one site has no relationship to a group using the same address at a different site. It also has no relationship to a non-permanent group using the same group ID with a different scope, or to a permanent group with the same group ID. Same address, totally different friend group 💅

Multicast addresses must not be used as source addresses in IPv6 packets or appear in any routing header. Hard rule, no exceptions 🚫💖

2.7.1. Pre-Defined Multicast Addresses ✨

okay so these are the multicast addresses that come already baked in, no setup needed 💖 here are the well-known ones:

Reserved Multicast Addresses:

FF00:0:0:0:0:0:0:0
FF01:0:0:0:0:0:0:0
FF02:0:0:0:0:0:0:0
FF03:0:0:0:0:0:0:0
FF04:0:0:0:0:0:0:0
FF05:0:0:0:0:0:0:0
FF06:0:0:0:0:0:0:0
FF07:0:0:0:0:0:0:0
FF08:0:0:0:0:0:0:0
FF09:0:0:0:0:0:0:0
FF0A:0:0:0:0:0:0:0
FF0B:0:0:0:0:0:0:0
FF0C:0:0:0:0:0:0:0
FF0D:0:0:0:0:0:0:0
FF0E:0:0:0:0:0:0:0
FF0F:0:0:0:0:0:0:0

these ones are reserved, and they SHALL NEVER be assigned to any multicast group 🙅‍♀️

All Nodes Addresses:

FF01:0:0:0:0:0:0:1
FF02:0:0:0:0:0:0:1

these identify the group of all IPv6 nodes, within scope 1 (node-local) or 2 (link-local) 🌸

All Routers Addresses:

FF01:0:0:0:0:0:0:2
FF02:0:0:0:0:0:0:2
FF05:0:0:0:0:0:0:2

these identify the group of all IPv6 routers, within scope 1 (node-local), 2 (link-local), or 5 (site-local) 👑

Solicited-Node Address:

FF02:0:0:0:0:1:FFXX:XXXX

ok this one is computed 🧚 it's a function of a node's unicast and anycast addresses. you take the low-order 24 bits of the address (unicast or anycast) and append those bits to the prefix FF02:0:0:0:0:1:FF00::/104, which gives you a multicast address somewhere in this range:

FF02:0:0:0:0:1:FF00:0000
to
FF02:0:0:0:0:1:FFFF:FFFF

for example, the solicited-node multicast address for the IPv6 address 4037::01:800:200E:8C6C is FF02::1:FF0E:8C6C 🦋 IPv6 addresses that differ only in the high-order bits (say, because there are multiple high-order prefixes tied to different aggregations) will all map to the same solicited-node address. that's lowkey genius, because it cuts down on how many multicast addresses a node has to join 🫶

a node is REQUIRED to compute and join the associated Solicited-Node multicast addresses for every unicast and anycast address it is assigned 💅

2.7.2. Assignment of New IPv6 Multicast Addresses ✨

okay the current approach [ETHER] for mapping IPv6 multicast addresses into IEEE 802 MAC addresses takes the low-order 32 bits of the IPv6 multicast address and uses them to build a MAC address 🎀 token ring networks are handled differently, and that's defined in [TOKEN]. group IDs less than or equal to 32 bits will generate unique MAC addresses, so new IPv6 multicast addresses SHOULD be assigned so that the group identifier is always in the low-order 32 bits, like this:

|   8    |  4 |  4 |          80 bits          |     32 bits     |
+------ -+----+----+---------------------------+-----------------+
|11111111|flgs|scop|   reserved must be zero   |    group ID     |
+--------+----+----+---------------------------+-----------------+

yes this limits the number of permanent IPv6 multicast groups to 2^32, but it's unlikely to be a limitation in the future 🌈 and if we ever DO need to go past that, multicast will still work, it'll just be slightly slower 🐢

additional IPv6 multicast addresses are defined and registered by the IANA [MASGN] 💜

2.8. A Node's Required Addresses ✨

a host is REQUIRED to recognize the following addresses as identifying itself:

  • Its Link-Local Address for each interface
  • Assigned Unicast Addresses
  • Loopback Address
  • All-Nodes Multicast Addresses
  • Solicited-Node Multicast Address for each of its assigned unicast and anycast addresses
  • Multicast Addresses of all other groups to which the host belongs.

a router is REQUIRED to recognize all of the addresses a host is REQUIRED to recognize, plus the following addresses as identifying itself:

  • The Subnet-Router anycast addresses for the interfaces it is configured to act as a router on.
  • All other Anycast addresses with which the router has been configured.
  • All-Routers Multicast Addresses
  • Multicast Addresses of all other groups to which the router belongs.

the only address prefixes which SHOULD be predefined in an implementation are the:

  • Unspecified Address
  • Loopback Address
  • Multicast Prefix (FF)
  • Local-Use Prefixes (Link-Local and Site-Local)
  • Pre-Defined Multicast Addresses
  • IPv4-Compatible Prefixes

implementations SHOULD assume all other addresses are unicast unless they're specifically configured (e.g., anycast addresses) 🦋

3. Security Considerations 🔒

okay bestie, this part is short but it matters 💀 IPv6 addressing documents do not have any direct impact on Internet infrastructure security. authentication of IPv6 packets is defined in [AUTH], so go read that one for the actual security stuff 🫶

APPENDIX A: Creating EUI-64 based Interface Identifiers 🧚

depending on the characteristics of a specific link or node, there are a number of approaches for creating EUI-64 based interface identifiers 💖 this appendix walks through some of them.

the only change needed to turn an EUI-64 identifier into an interface identifier is to invert the "u" (universal/local) bit ✨ for example, a globally unique EUI-64 identifier looks like this:

|0              1|1              3|3              4|4              6|
|0              5|6              1|2              7|8              3|
+----------------+----------------+----------------+----------------+
|cccccc0gcccccccc|ccccccccmmmmmmmm|mmmmmmmmmmmmmmmm|mmmmmmmmmmmmmmmm|
+----------------+----------------+----------------+----------------+

where "c" are the bits of the assigned company_id, "0" is the value of the universal/local bit to indicate global scope, "g" is individual/group bit, and "m" are the bits of the manufacturer-selected extension identifier. the IPv6 interface identifier would then look like this:

|0              1|1              3|3              4|4              6|
|0              5|6              1|2              7|8              3|
+----------------+----------------+----------------+----------------+
|cccccc1gcccccccc|ccccccccmmmmmmmm|mmmmmmmmmmmmmmmm|mmmmmmmmmmmmmmmm|
+----------------+----------------+----------------+----------------+

the only change is inverting the value of the universal/local bit 🔥

[EUI64] defines a method to create an EUI-64 identifier from an IEEE 48bit MAC identifier 🌸 you insert two octets, with hexadecimal values of 0xFF and 0xFE, in the middle of the 48 bit MAC (between the company_id and vendor supplied id). for example, the 48 bit MAC with global scope looks like this:

|0              1|1              3|3              4|
|0              5|6              1|2              7|
+----------------+----------------+----------------+
|cccccc0gcccccccc|ccccccccmmmmmmmm|mmmmmmmmmmmmmmmm|
+----------------+----------------+----------------+

where "c" are the bits of the assigned company_id, "0" is the value of the universal/local bit to indicate global scope, "g" is individual/group bit, and "m" are the bits of the manufacturer-selected extension identifier. the interface identifier would be of the form:

|0              1|1              3|3              4|4              6|
|0              5|6              1|2              7|8              3|
+----------------+----------------+----------------+----------------+
|cccccc1gcccccccc|cccccccc11111111|11111110mmmmmmmm|mmmmmmmmmmmmmmmm|
+----------------+----------------+----------------+----------------+

when IEEE 802 48bit MAC addresses are available (on an interface or a node), an implementation SHOULD use them to create interface identifiers, because of their availability and uniqueness properties 💜

there are a number of link types that are multi-access but don't have globally unique link identifiers 🦋 examples include LocalTalk and Arcnet. the way to create an EUI-64 formatted identifier here is to take the link identifier (e.g., the LocalTalk 8 bit node identifier) and zero fill it to the left. for example, a LocalTalk 8 bit node identifier of hexadecimal value 0x4F results in this interface identifier:

|0              1|1              3|3              4|4              6|
|0              5|6              1|2              7|8              3|
+----------------+----------------+----------------+----------------+
|0000000000000000|0000000000000000|0000000000000000|0000000001001111|
+----------------+----------------+----------------+----------------+

note that this results in the universal/local bit set to "0" to indicate local scope 🌈

there are a bunch of links that don't have any built-in identifier. the most common ones are serial links and configured tunnels 🥺 so interface identifiers MUST be chosen that are unique for the link.

when there's no built-in identifier available on a link, the preferred approach is to use a global interface identifier from another interface, or one that's assigned to the node itself. to use this approach, no other interface connecting the same node to the same link MAY use the same identifier 🙈

if there's no global interface identifier available for use on the link, the implementation needs to create a local scope interface identifier. the only requirement is that it be unique on the link. there are many possible approaches for picking a link-unique interface identifier, including:

  • Manual Configuration
  • Generated Random Number
  • Node Serial Number (or other node-specific token)

the link-unique interface identifier SHOULD be generated in a way that it doesn't change after a reboot of a node or if interfaces are added or deleted from the node 🔁

which algorithm you pick depends on the link and the implementation 🍓 the details on forming interface identifiers are defined in the appropriate "IPv6 over <link>" specification. it is strongly RECOMMENDED that a collision detection algorithm be implemented as part of any automatic algorithm 💅

APPENDIX B: ABNF Description of Text Representations 🎀

this appendix defines the text representation of IPv6 addresses and prefixes in Augmented BNF [ABNF] for reference purposes 💖

   IPv6address = hexpart [ ":" IPv4address ]
   IPv4address = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT

   IPv6prefix  = hexpart "/" 1*2DIGIT

   hexpart = hexseq | hexseq "::" [ hexseq ] | "::" [ hexseq ]
   hexseq  = hex4 *( ":" hex4)
   hex4    = 1*4HEXDIG

APPENDIX C: CHANGES FROM RFC-1884 🌈

Okay bestie, here's the full glow-up list of everything that changed from RFC 1884 💖

  • Added an appendix with an ABNF description of text representations 📝
  • Clarified that link unique identifiers do not change after a reboot or other interface reconfigurations 🔒
  • Clarified the Address Model based on comments 💬
  • Changed the aggregation format terminology so it matches the aggregation draft 🧩
  • Added text that lets an interface identifier be used on more than one interface on the same node 🦋
  • Added rules for defining new multicast addresses 📮
  • Added an appendix describing how to create EUI-64 based interface IDs 🌸
  • Added notation for defining IPv6 prefixes ✨
  • Changed the solicited-node multicast definition to use a longer prefix 📏
  • Added the site-scope all-routers multicast address 🗺️
  • Defined Aggregatable Global Unicast Addresses to use the "001" Format Prefix 🎀
  • Changed the "010" (Provider-Based Unicast) and "100" (Reserved for Geographic) Format Prefixes to Unassigned 🙈
  • Added a section on interface ID definition for unicast addresses. It requires EUI-64 to be used within the range of format prefixes, and it sets rules for the global/local scope bit in EUI-64 🌈
  • Updated the NSAP text to reflect the work in RFC1888 🌷
  • Removed protocol-specific IPv6 multicast addresses (for example, DHCP) and referenced the IANA definitions instead 🧚
  • Removed the section "Unicast Address Example". It had become OBE (overtaken by events) 💀
  • Added new and updated references 📚
  • Minor text clarifications and improvements 💅

REFERENCES 📚

[ABNF] Crocker, D., and P. Overell, "Augmented BNF for Syntax Specifications: ABNF", RFC 2234, November 1997. This one is the grammar language we've been using all along 🫶

[AGGR] Hinden, R., O'Dell, M., and S. Deering, "An Aggregatable Global Unicast Address Format", RFC 2374, July 1998. The aggregation glow-up 💅

[AUTH] Atkinson, R., "IP Authentication Header", RFC 1826, August 1995. 🔐

[ANYCST] Partridge, C., Mendez, T., and W. Milliken, "Host Anycasting Service", RFC 1546, November 1993. 🌍

[CIDR] Fuller, V., Li, T., Yu, J., and K. Varadhan, "Classless Inter-Domain Routing (CIDR): An Address Assignment and Aggregation Strategy", RFC 1519, September 1993. The OG address-splitting queen 👑

[ETHER] Crawford, M., "Transmission of IPv6 Pacekts over Ethernet Networks", Work in Progress. 🧵

[EUI64] IEEE, "Guidelines for 64-bit Global Identifier (EUI-64) Registration Authority", http://standards.ieee.org/db/oui/tutorials/EUI64.html, March 1997. The EUI-64 source of truth 🆔

[FDDI] Crawford, M., "Transmission of IPv6 Packets over FDDI Networks", Work in Progress. 💫

[IPV6] Deering, S., and R. Hinden, Editors, "Internet Protocol, Version 6 (IPv6) Specification", RFC 1883, December 1995. The original IPv6 spec, iconic 🌟

[MASGN] Hinden, R., and S. Deering, "IPv6 Multicast Address Assignments", RFC 2375, July 1998. 📮

[NSAP] Bound, J., Carpenter, B., Harrington, D., Houldsworth, J., and A. Lloyd, "OSI NSAPs and IPv6", RFC 1888, August 1996. 🌐

[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. The keyword rulebook 📖

[TOKEN] Thomas, S., "Transmission of IPv6 Packets over Token Ring Networks", Work in Progress. 🪢

[TRAN] Gilligan, R., and E. Nordmark, "Transition Mechanisms for IPv6 Hosts and Routers", RFC 1993, April 1996. Transition era vibes 🦋

AUTHORS' ADDRESSES 💌

Robert M. Hinden
Nokia
232 Java Drive
Sunnyvale, CA 94089
USA

Phone: +1 408 990-2004
Fax:   +1 408 743-5677
EMail: hinden@iprg.nokia.com


Stephen E. Deering
Cisco Systems, Inc.
170 West Tasman Drive
San Jose, CA 95134-1706
USA

Phone: +1 408 527-8213
Fax:   +1 408 527-8254
EMail: deering@cisco.com

This part is legal text, so I've kept the wording exactly as written so the permissions still mean what they mean 🫶

Copyright (C) The Internet Society (1998). All Rights Reserved.

This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English.

The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns.

This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. That's the whole thing, bestie 🌈✨

📦 where this page comes from