Internet Protocol version 6 (IPv6) is an Internet Layer protocol for packet-switched internetworks. IPv4 is currently the dominant Internet Protocol version, and was the first to receive widespread use. The Internet Engineering Task Force (IETF) has designated IPv6 as the successor to version 4 for general use on the Internet.
IPv6 has a much larger address space than IPv4, which provides flexibility in allocating addresses and routing traffic. The extended address length (128 bits) is intended to eliminate the need for network address translation to avoid address exhaustion, and also simplifies aspects of address assignment and renumbering, when changing Internet connectivity providers.
The very large IPv6 address space supports 2128 (about 3.4×1038) addresses, or approximately 5×1028 (roughly 295) addresses for each of the roughly 6.5 billion (6.5×109) people alive today. In a different perspective, this is 252 addresses for every observable star in the known universe – more than ten billion billion billion times as many addresses as IPv4 (232) supported.
While these numbers are impressive, it was not the intent of the designers of the IPv6 address space to assure geographical saturation with usable addresses. Rather, the large number allows a better, systematic, hierarchical allocation of addresses and efficient route aggregation. With IPv4, complex Classless Inter-Domain Routing (CIDR) techniques were developed to make the best use of the small address space. Renumbering an existing network for a new connectivity provider with different routing prefixes is a major effort with IPv4, as discussed in RFC 2071 and RFC 2072. With IPv6, however, changing the prefix in a few routers can renumber an entire network ad hoc, because the host identifiers (the least-significant 64 bits of an address) are decoupled from the subnet identifiers and the network provider's routing prefix. The size of each subnet in IPv6 is 264 addresses (64 bits); the square of the size of the entire IPv4 Internet. Thus, actual address space utilization rates will likely be small in IPv6, but network management and routing will be more efficient.
During the first decade of operation of the TCP/IP-based Internet, by the late 1980s, it became apparent that methods had to be developed to conserve address space. In the early 1990s, even after the introduction of classless network redesign, it was clear that this was not enough to prevent IPv4 address exhaustion and that further changes to the Internet infrastructure were needed. By the beginning of 1992, several proposed systems were being circulated, and by the end of 1992, the IETF announced a call for white papers (RFC 1550) and the creation of the "IP Next Generation" (IPng) area of working groups.
The Internet Engineering Task Force adopted IPng on July 25, 1994, with the formation of several IPng working groups. By 1996, a series of RFCs were released defining Internet Protocol Version 6 (IPv6), starting with RFC 2460.
Incidentally, the IPng architects could not use version number 5 as a successor to IPv4, because it had been assigned to an experimental flow-oriented streaming protocol (Internet Stream Protocol), similar to IPv4, intended to support video and audio.
It is widely expected that IPv4 will be supported alongside IPv6 for the foreseeable future. IPv4-only nodes are not able to communicate directly with IPv6 nodes, and will need assistance from an intermediary; see Transition mechanisms, below.
IPv6 specifies a new packet format, designed to minimize packet-header processing. Since the headers of IPv4 and IPv6 are significantly different, the two protocols are not interoperable.
Interfaces configured for IPv6 almost always have more than one address, usually one for the local link (the link-local address), and additional ones for site-local or global addressing. Link-local addresses are often used in network address autoconfiguration where no external source of network addressing information is available.
In addition to address scopes, IPv6 introduces the concept of "scope zones". Each address can only belong to one zone corresponding to its scope. A "link zone" (link-local zone) consists of all network interfaces connected on one link. Addresses maintain their uniqueness only inside a given scope zone. Zones are indicated by a suffix (zone index) to an address. For example, fe80::211:d800:97:c915%eth0 (link-local address) and fec0:0:0:ffff::1%4 (site-local address) show the additional suffix indicated by the percent (%) character.
If IPv6 stateless address autoconfiguration (SLAAC) proves unsuitable, a host can use stateful configuration (DHCPv6) or be configured manually. In particular, stateless autoconfiguration is not used by routers, these must be configured manually or by other means.
IPv6 does not implement broadcast, the ability to send a packet to all hosts on the attached link. The same effect can be achieved by sending a packet to the link-local all hosts multicast group.
Most environments, however, do not currently have their network infrastructures configured to route multicast packets; multicasting on single subnet will work, but global multicasting might not.
IPv4 maintained a checksum packet header field that covers the entire packet header. Since certain fields (such as the TTL field) change during forwarding, every router must re-compute the checksum. IPv6 has no error checking at the Internet Layer, but instead relies on Link Layer and transport protocols to perform error checking.
As checksum computation in modern backbone routers is usually performed in hardware at link speed, performance gains based on eliminated checksums might be marginal in IPv6.
IPv6 routers do not handle packet fragmentation. If necessary, this is also delegated to the communication end points.
The IPv6 packet is composed of two main parts: the header and the payload.
The header is in the first 40 octets (320 bits) of the packet and contains:
The payload can have a size of up to 64KiB in standard mode, or larger with a "jumbo payload" option.
The protocol field of IPv4 is replaced with a Next Header field. This field usually specifies the transport layer protocol used by a packet's payload.
In the presence of options, however, the Next Header field specifies the presence of an extra options header, which then follows the IPv6 header; the payload's protocol itself is specified in a field of the options header. This insertion of an extra header to carry options is analogous to the handling of AH and ESP in IPsec for both IPv4 and IPv6.
The length of network addresses emphasize a most important change when moving from IPv4 to IPv6. IPv6 addresses are 128 bits long (as defined by RFC 4291), whereas IPv4 addresses are 32 bits; where the IPv4 address space contains roughly 4 billion addresses, IPv6 has enough room for 3.4×1038 unique addresses.
IPv6 addresses are typically composed of two logical parts: a 64-bit (sub-)network prefix, and a 64-bit host part, which is either automatically generated from the interface's MAC address or assigned sequentially. Because the globally unique MAC addresses offer an opportunity to track user equipment, and so users, across time and IPv6 address changes, RFC 3041 was developed to reduce the prospect of user identity being permanently tied to an IPv6 address, thus restoring some of the possibilities of anonymity existing at IPv4. RFC 3041 specifies a mechanism by which time-varying random bit strings can be used as interface circuit identifiers, replacing unchanging and traceable MAC addresses.
IPv6 addresses are normally written as eight groups of four hexadecimal digits, where each group is separated by a colon (:). For example,
is a valid IPv6 address. To shorten the writing and presentation of addresses, several simplifications to the notation are permitted.
Any leading zeros in a group may be omitted; thus, the given example becomes
One or any number of consecutive groups of 0 value may be replaced with two colons (::):
This substitution with double-colon may be performed only once in an address, because multiple occurrences would lead to ambiguity. For example, the address 2001:0:0:FFD3:0:0:0:57ab, if written as 2001::FFD3::57ab, could represent 2001:0:0:0:0:FFD3:0:57ab, 2001:0:0:0:FFD3:0:0:57ab, 2001:0:0:FFD3:0:0:0:57ab, and 2001:0:FFD3:0:0:0:0:57ab.
Accordingly, the localhost (loopback) address, fully written as 0000:0000:0000:0000:0000:0000:0000:0001, may be reduced to ::1 and the undetermined IPv6 address (zero value), i.e., all bits are zero, is simply ::.
For comparison, the addresses below are all valid and equivalent:
In a special group of IPv6 addresses called "compatible addresses" (see below), the sequence of the last 4 bytes of the IPv6 address may be written in dot-decimal notation, in the style of IPv4 addresses. This notation is convenient when working in a mixed (dual-stack) environment of IPv4 and IPv6 addresses. The general form of the notation is x:x:x:x:x:x:d.d.d.d, where the x's are the 6 high-order groups of hexadecimal digits and the d's represent the decimal digit groups of the four low-order octets of the address. For example, ::ffff:126.96.36.199 is the same address as ::ffff:0c22:384e. Usage of this notation may not be widely supported.
RFC 4291 (IP Version 6 Addressing Architecture) provides additional information.
Such blocks of consecutive IPv6 addresses are written using the same notation previously developed for IPv4 Classless Inter-Domain Routing (CIDR). CIDR notation designates a leading set of bits by appending the size (in decimal) of that bit block (prefix) to the address, separated by a forward slash character (/).
For example, a network is denoted by the first address in the network and the bit block size of the prefix, such as 2001:0db8:1234::/48. The network starts at address 2001:0db8:1234:0000:0000:0000:0000:0000 and ends at 2001:0db8:1234:ffff:ffff:ffff:ffff:ffff.
Single host addresses are often also written in CIDR notation to indicate the routing behavior of the network they belong to. For example, the address 2001:db8:a::123/128 indicates a single interface route for this address, whereas 2001:db8:a::123/32 may indicate a different routing environment.
There are a number of addresses with special meaning in IPv6:Unspecified address
For example, host A has two interfaces which automatically receive link-local addresses when activated (per RFC 4862), say, fe80::1/64 and fe80::2/64. Only one of these is connected to the same physical network as another host B which has address fe80::3/64. If host A attempts to contact fe80::3 it cannot a priori know which interface (fe80::1 or fe80::2) to use based on the prefix route alone.
For this distinction, RFC 4007 provides the definition of address scopes. An IP address can only be unique within a given scope. The address scope is expressed by addition of a unique zone index for each interface, represented syntactically in the form %
Zone index notations are implemented differently in various operating systems:
Relatively few IPv6-capable applications understand address scope syntax at the user level, thus rendering link-local addressing inappropriate for many user applications. However, link-local addresses are not intended for most of such application usage and their primary benefit is in low-level network management functions.
Optionally, special hosts may require additional multicast addresses of the appropriate type, e.g., a router needs support for the all-routers multicast address.
Example: In a URL the IPv6-Address is enclosed in brackets, e.g.,
If the URL also contains a port number the notation is:
This is not only useful but mandated when using shortform:
Additional information can be found in "RFC 2732 - Format for Literal IPv6 Addresses in URL's" and "RFC 3986 - Uniform Resource Identifier (URI): Generic Syntax."
In Microsoft Windows operating systems, IP addresses were also allowed in Uniform Naming Convention (UNC) path names. Since the colon is an illegal character in a UNC path name, the use of IPv6 addresses is also illegal in UNC names. For this reason, Microsoft has registered a second-level Internet domain, ipv6-literal.net, as a means to facilitate symbolic substitution. IPv6 addresses may be transcribed in the following fashion:
is written as
This notation is automatically resolved by Microsoft software without DNS queries to any nameservers. If the IPv6 address contains a zone index, it is appended to the address portion after an 's' character:
At the design-stage of the IPv6 DNS architecture, the AAAA scheme faced a rival proposal. This alternate approach, designed to facilitate network renumbering, uses A6 records for the forward lookup and a number of other innovations such as bit-string labels and DNAME records. It is defined in RFC 2874 and its references (with further discussion of the pros and cons of both schemes in RFC 3364), but has been deprecated to experimental status.
|TTL||Time to live in seconds|
|RDLENGTH||Length of RDATA field|
|RDATA||String form of the IPV6 address as described in RFC 3513|
RFC 3484 specifies how applications should select an IPv6 or IPv4 address for use, including addresses retrieved from DNS.
Until IPv6 completely supplants IPv4, a number of so-called transition mechanisms are needed to enable IPv6-only hosts to reach IPv4 services and to allow isolated IPv6 hosts and networks to reach the IPv6 Internet over the IPv4 infrastructure.
For the period while IPv6 hosts and routers co-exist with IPv4 systems, RFC 2893 (Transition Mechanisms for IPv6 Hosts and Routers) and RFC2185 (Routing Aspects of IPv6 Transition) define compatibility and transition mechanisms. These techniques, sometimes collectively called Simple Internet Transition (SIT), include:
Most current implementations of IPv6 use a dual stack. Some early experimental implementations used independent IPv4 and IPv6 stacks.
The direct encapsulation of IPv6 datagrams within IPv4 packets is indicated by IP protocol number 41. IPv6 can also be encapsulated within UDP packets e.g. in order to cross a router or NAT device that blocks protocol 41 traffic. Other encapsulation schemes, such as used in AYIYA or GRE, are also popular.
Another automatic tunneling mechanism, the ISATAP protocol, treats the IPv4 network as a virtual IPv6 local link, with mappings from each IPv4 address to a link-local IPv6 address.
Teredo, an automatic tunneling technique that uses UDP encapsulation, can allegedly cross multiple NAT boxes. Teredo is not widely deployed today, but an experimental version of Teredo is installed with the Windows XP SP2 IPv6 stack. IPv6, 6to4 and Teredo are enabled by default in Windows Vista and Mac OS X Leopard and Apple's AirPort Extreme.
Configured tunneling uses protocol 41 in the Protocol field of the IPv4 packet. This method is also known as 6in4.
After the Regional Internet Registries have exhausted their pools of available IPv4 addresses, it is likely that hosts newly added to the Internet, might only have IPv6 connectivity. For these clients to have backward-compatible connectivity to existing IPv4-only resources, suitable translation mechanisms must be deployed.
One form of translation is the use of a dual-stack application-layer proxy, for example a web proxy.
NAT-like techniques for application-agnostic translation at the lower layers have also been proposed. Most have been found to be too unreliable in practice because of the wide range of functionality required by common application-layer protocols, and are considered by many to be obsolete.
This "slow-down" results from DNS resolution failures due to faulty NAT 'routers' and other DNS resolvers which improperly handle the AAAA DNS query. These DNS resolvers just drop the DNS request for AAAA records, instead of properly returning the appropriate negative DNS response. Because the request is dropped, the host sending the request has to wait for a timeout to happen, thus causing a perceived slow down when connecting to new hosts. Since there is no result of the request that could be cached locally, even if a DNS cache is running, the problem will persist for identical lookups in the future.
Note that DNS queries may happen over any transport available (IPv4, if the only protocol), as the transport protocol used for a query is independent of the type of query.
A few international organizations are involved with IPv6 test and evaluation ranging from the United States Department of Defense to the University of New Hampshire.
Estimates as to when the pool of available IPv4 addresses will be exhausted used to vary widely. In 2003, Paul Wilson (director of APNIC) stated that, based on then-current rates of deployment, the available space would last until 2023. In September 2005 a report by Cisco Systems (a network hardware manufacturer) suggested that the pool of available addresses would dry up in as little as 4 to 5 years. As of November 2007, a daily updated report projected that the IANA pool of unallocated addresses would be exhausted in May 2010, with the various Regional Internet Registries using up their allocations from IANA in April 2011. There is now consensus among Regional Internet Registries that significant milestones of the exhaustion process will be met in 2010 or 2011, at the latest, and a policy process has started for the end-game and post-exhaustion era .
When the RIR and IANA pools are exhausted, there will still be unused IPv4 addresses, however, the existing mechanisms for allocating those addresses would no longer work. Mechanisms that have been discussed for allocating IPv4 addresses beyond this point have included the reclamation of unused address space, re-engineering hosts and routers to allow the use of areas of the IPv4 address space which are currently unusable for technical reasons, and the creation of a market in IPv4 addresses.
There are two distinct classes of users of networking equipment, informed (mainly commercial and professional), and uninformed (mainly consumer). The former understand that network devices are specialist computers which may need software upgrades for security and performance fixes. The latter generally treat their networking equipment as appliances, which are configured only when first unboxed, if at all, and only ever undergo firmware upgrades when absolutely necessary. Inevitably it is the latter group who have no knowledge of IPv4 or v6, but who are most likely to suffer when their equipment has to be replaced, since commercial grade equipment has generally handled IPv6 for quite a few years.
Most equipment such as hosts and routers require explicit IPv6 support. Fewer problems arise with equipment which only does low-level transport, such as cables, most ethernet adapters, and most layer-2 switches.
As of 2007, IPv6 readiness is currently not considered in most consumer purchasing decisions. If such equipment is not IPv6-capable, it might need to be upgraded or replaced prematurely if connectivity from or to new users and to servers using IPv6 addresses is required.
As with the year-2000 compatibility, IPv6 compatibility is mainly a software/firmware issue. However, unlike the year-2000 issue, there seems to be virtually no effort to ensure compatibility of older equipment and software by manufacturers. Furthermore, even compatibility of products now available is unlikely for many types of software and equipment. This is caused by only a recent realisation that IPv4 exhaustion is imminent, and the hope that we will be able to get by for a relatively long time with a combined IPv4/IPv6 situation. There is a tug-of-war going on in the internet community whether the transition will/should be rapid or long. Specifically, an important question is whether almost all internet servers should be ready to serve to new IPv6-only clients by 2012. Universal access to IPv6-only servers will be even more of a challenge.
Most equipment would be fully IPv6 capable with a software/firmware update if the device has sufficient code and data space to support the additional protocol stack. However, as with 64-bit Windows and Wi-Fi Protected Access support, manufacturers are likely to try to save on development costs for hardware which they no longer sell, and to try to get more sales from new "IPv6-ready" equipment. Even when chipset makers develop new drivers for their chipsets, device manufacturers might not pass these on to the consumers. Moreover, as IPv6 gets implemented, optional features might become really important, such as IPv6 mobile. It is therefore important to check your supplier on its support record, and get guarantees if you can or need to. Examples of equipment which currently usually are not IPv6 ready, are home routers. As for the CableLabs consortium, the 160 Mbit/s DOCSIS 3.0 IPv6-ready specification for cable modems has only been issued in August 2006. IPv6 capable Docsis 2.0b was skipped while the widely used DOCSIS 2.0 does not support IPv6. The new 'DOCSIS 2.0 + IPv6' standard also supports IPv6, which may on the cable modem side only require a firmware upgrade . It is expected that only 60% of cable modems' servers and 40% of cable modems will be DOCSIS 3.0 by 2011. Other equipment which is typically not IPv6-ready range from Skype and SIP phones to oscilloscopes and printers. Professional network routers in use should be IPv6-ready. Most personal computers should also be IPv6-ready, because the network stack resides in the operating system. Most applications with network capabilities are not ready, but could be upgraded with support from the developers. Since February 2002, with J2SE 1.4, all applications that are 100% Java have implicit support for IPv6 addresses.
ADSL services offer a problem if the access networks of the incumbent telephone connection cannot support IPv6, such that independent ADSL providers cannot provide native IPv6 connectivity.
Although IPv4 address exhaustion has been slowed by the introduction of classless inter-domain routing (CIDR) and the extensive use of network address translation (NAT), address uptake has accelerated again in recent years. Some forecasts expect complete depletion by the year 2011.
As of 2008, IPv6 accounts for a minuscule fraction of the used addresses and the traffic in the publicly-accessible Internet which is still dominated by IPv4.
The 2008 Summer Olympic Games were a notable event in terms of IPv6 deployment. For the first time a major World event has had a presence on the IPv6 Internet at http://ipv6.beijing2008.cn/en (IP addresses 2001:252:0:1::2008:6 and 2001:252:0:1::2008:8) and all network operations of the Games were conducted using IPv6.It is believed that the Olympics provided the largest showcase of IPv6 technology since the inception of IPv6.
|Year||Announcements and availability|
|1996||Alpha quality IPv6 support in Linux kernel development version 2.1.8.|
|1997||By the end of 1997, a large number of interoperable implementations existed.|
|In the end of 1997 IBM's AIX 4.3 was the first commercial platform that supported IPv6.|
|1998||Microsoft Research first released an experimental IPv6 stack in 1998. This support was not intended for use in a production environment.|
|2000||Production-quality BSD support for IPv6 became generally available in early to mid-2000 in FreeBSD, OpenBSD, and NetBSD via the KAME project.|
|Microsoft releases an IPv6 technology preview version for Windows 2000 in March 2000.|
|Sun Solaris has IPv6 support since Solaris 8 in February 2000.|
|2001||Cisco Systems introduced IPv6 support on Cisco IOS routers and L3 switches in 2001.|
|2002||Microsoft Windows NT 4.0 and Windows 2000 SP1 had limited IPv6 support for research and testing since at least 2002.|
|Microsoft Windows XP (2001) had IPv6 support for developmental purposes. In Windows XP SP1 (2002) and Windows Server 2003, IPv6 is included as a core networking technology, suitable for commercial deployment.|
|IBM z/OS has supported IPv6 since version 1.4, which became generally available in September 2002.|
|2003||Apple Mac OS X v10.3 "Panther" (2003) has IPv6 supported and enabled by default.|
|In July, ICANN announced that the IPv6 AAAA records for the Japan (.jp) and Korea (.kr) country code Top Level Domain (ccTLD) nameservers became visible in the DNS root server zone files with serial number 2004072000. The IPv6 records for France (.fr) were added a little later. This made IPv6 operational in a public fashion.|
|2005||Linux 2.6.12 removes IPv6's "experimental" status.|
|2007||Microsoft Windows Vista (2007) has IPv6 supported and enabled by default.|
|Apple's AirPort Extreme 802.11n base station consists of an IPv6 gateway in its default configuration. It uses 6to4 tunneling and can optionally route through a manually configured IPv4 tunnel.|
|2008||On February 4, 2008, IANA added AAAA records for the IPv6 addresses of six of the thirteen root name servers. With this transition, it is now possible for two Internet hosts to communicate via DNS without using IPv4 at all.|
|On March 12, 2008, Google launched a public IPv6 web interface to its popular search engine at the URL http://ipv6.google.com|