IPv4 address exhaustion

Article on other languages:

• Agotamiento_de_las_direcciones_IPv4
• IPアドレス枯渇問題

	del.icio.us
	Digg
	Furl
	Reddit
	Rojo
	Add to OnlyWire

The neutrality of this article is disputed. Please see the discussion on the talk page.(December 2007) Please do not remove this message until the dispute is resolved.

Internet addressing growth map.

IP address exhaustion refers to the decreasing availability of publicly available IPv4 IP addresses. This has been a concern that has spanned decades since the 1980s. As a result, this has been the driving factor in creating and adopting several new technologies, including classful networks, CIDR addressing, Network Address Translation (NAT) and IPv6.

As of 2008, IPv6 is generally seen as the only practical long-term solution for IPv4 address exhaustion that is readily available. However, although the predicted IPv4 address exhaustion approaches, most ISPs, software vendors and service providers are only just starting to consider widespread deployment of IPv6.

Synopsis

Main article: IP Address

Every host on a network, such as a computer or networked printer, is assigned a unique IP address that is used to communicate with other hosts on that network, normally expressed in dotted decimal format (for example 66.230.200.110). Each octet, or part of the address, must be a number from 0 to 255 and therefore there is a logical maximum of 4,294,967,296 addresses available for use. However, large numbers of addresses are reserved for local use and are unavailable for Internet use.

There are insufficient publicly routable IPv4 addresses to provide a distinct address to every IPv4 speaker (which include desktop computers, mobile phones, embedded devices, and virtual hosts). This problem is mitigated by network address translation (NAT), whereby a single public Internet IP address can be shared by multiple internal local area network (LAN) hosts. Data sent by individual hosts to the Internet states its source address as the public IP address used, and the router providing the access is able to keep track of which host originated the traffic inside the network and forwards replies accordingly. This is similar to multiple office telephones that share one phone number by using extension numbers to distinguish individual telephones.

Exhaustion date

This article or section may contain an inappropriate mixture of prose and timeline. Please help convert this timeline into prose or, if necessary, a list.

Exhaustion will occur on all continents at the same time, as all registries follow similar allocation policies, with for about 12 to 18 months stock allocated at each request. Only specific organisations which requested addresses in the pre-CIDR or pre-RIR eras possibly have a significant stock left.

• As of June 2008, Geoff Huston of APNIC predicts with detailed simulations an exhaustion of the unallocated IANA pool in January 2011.^[1] Tony Hain of networking equipment manufacturer Cisco Systems predicts the exhaustion date to be around October 2010.^[2] These predictions are derived from current trends, and do not take into account any last chance rush to acquire the last available addresses. After the IANA pool exhaustion, during 11 months each individual regional Internet registry (RIR) will be able to supply with their last assigned addresses. These dates lie within a depreciation time of five to ten years of network equipment that is currently being acquired.
• On May 21, 2007, the American Registry for Internet Numbers (ARIN), the North American RIR, advised the Internet community that due to the expected exhaustion in 2010 "migration to IPv6 numbering resources is necessary for any applications which require ongoing availability from ARIN of contiguous IP numbering resources".^[3] It should be noted that "applications" include general connectivity between devices on the Internet, as some devices only have an IPv6 address allocated.
• On June 20, 2007, the Latin American and Caribbean Internet Addresses Registry (LACNIC), the South American RIR, advised "preparing its regional networks for IPv6" by January 1, 2011 for the exhaustion of IPv4 addresses "in three years time".^[4]
• On June 26, 2007, the Asia-Pacific Network Information Centre (APNIC), the RIR for the Pacific and Asia, endorsed a statement by the Japan Network Information Center (JPNIC) that to continue the expansion and development of the Internet a move towards an IPv6-based Internet is advised. This with an eye on the expected exhaustion around 2010 which will create a great restriction on the Internet.^[5][6]

Less than four years until the first RIR exhaustion is a short time for the entire industry to transition to IPv6. This situation is aggravated by the fact that until exhaustion there will be no significant demand. David Conrad, the general manager of IANA acknowledges, "I suspect we are actually beyond a reasonable time frame where there won't be some disruption. Now it's more a question of how much." Geoff Huston claims we should have started the transition to IPv6 much earlier, such that by the exhaustion date it would be completed, with all devices IPv6-capable, and IPv4 getting phased out.

It should be recognized that by the end of 2011, there will be new clients and servers on the Internet which have no choice but to only have an IPv6 address. For the rest of the Internet to be able to communicate with them they should then be able to a) serve to IPv6 customers, and b) to access IPv6 servers. Within scalable solutions, the first requires Internet-facing servers to be on IPv6, and the second requires pretty much all devices to be on IPv6.

Causes

Several forces threaten the Internet with address exhaustion. Each of them drastically increases the demand on the limited supply of 32-bit addresses, often in ways unanticipated by the original designers of the network.

Mobile devices

Just as IPv4 has become the de facto standard for networked communication, the cost of embedding substantial computing power into handheld devices has plummeted. As a result, formerly "dumb" mass-market devices such as mobile phones have become potential Internet hosts. With mobile phone market penetration approaching 100% across the world, the result is a plausible scenario in which every person on the planet could be IP-addressable.^[7]

Always-on connections

Throughout the 1990s, the predominant mode of consumer Internet access was dialup Internet access. Dialup access reduces pressure on IP addresses, because dialup links are usually disconnected and thus do not require permanent IP addresses. By 2007, however, broadband Internet access had begun to exceed 50% penetration in many markets. Broadband connections remain constantly active, and even when dynamically addressed, still require a persistent IP address.^[8]

Internet demographics

There are hundreds of millions of households in the developed world. In 1990, only a bare fraction of these had Internet connectivity. Just 15 years later, almost half of them had persistent broadband connections.^[9]

Inefficient address use

Organizations that obtained IP addresses in the 1980s were often allocated far more addresses than they actually required. For example, large companies or universities were given class A address blocks, each of which contained 16 million IPv4 addresses. Many organisations continue to utilise public IP addresses for devices not accessible outside their local network and would be well served by a NAT based implementation, releasing potentially large ranges of IP addresses for re-allocation. Some organisations also have large ranges of IP addresses currently not utilised but which have not been released back to the allocation authorities for various reasons.

Due to inefficiencies caused by subnetting, it is very difficult to use all the addresses in a block. The Host-Density ratio, as defined in RFC 3194, is an intuitive metric for utilization of IP address blocks.

Mitigation

Some things that can be done to mitigate the IPv4 address exhaustion are (not mutually exclusive):

• Network address translation (NAT)
• Use of private networks
• Dynamic Host Configuration Protocol (DHCP)
• Name-based virtual hosting
• Tighter control by regional Internet registries on the allocation of addresses to local Internet registries
• Network renumbering to reclaim large blocks of address space allocated in the early days of the Internet

Conservation

"Conservation" is another method used to preserve available IP addresses. Upon conception of the Internet it was never envisaged that it would require anywhere near as many IP addresses as it now does; therefore they were frequently allocated in 'blocks' of 255, 65536, or 16777216 addresses for use. To this day several organisations have been assigned 16 million IP addresses of which they use a comparative handful. These days organisations responsible for allocation of public IP addresses are much more reluctant to assign large groups.

Subnetting

Subnetting is again another method used to get more use out of IP addresses generally, in short the dotted decimal notation is a user-friendly method of representing binary addresses such as 01000010111001101100100001101110 (again 66.230.200.110). These addresses are subnetted by applying a subnet mask which denotes which portion of the address is the network portion and which is the host portion; this is analogous to the area code and subscriber number of a telephone number, the phone number (212) 555-9293 is uniquely identifiable from (213) 555-9293. This allows the same numbers to be used in multiple locations with only some minor extra consideration.

Reclaiming unused IPv4 space

The neutrality or factuality of this article or section may be compromised by weasel words, which can allow the implication of unsourced information. You can help Wikipedia by removing weasel worded statements.

In the early days of the Internet, before the creation of classful networks and later CIDR addressing, large blocks of IP addresses were allocated to individual companies and organizations. IANA could potentially reclaim these ranges and reissue the addresses to others. However, it can cost a great deal of time and money to renumber a network so these organizations will likely object, quite possibly to the point of filing lawsuits. Moreover, at the current rate of IPv4 address consumption, even if all of these could be reclaimed, it would result in only extending the address exhaustion date back a year or two.^{[citation needed]}

Similarly, many IP address have been allocated to companies that no longer exist or were never used. Unfortunately, the stricter accounting of IP address allocation currently in place was not always in place and it would take quite a bit of effort to track down which addresses really are unused. Many IP addresses that do not show up in the public BGP routing tables are actually in use on intranets. Again, it is likely that more time would be spent tracking down which IP address could be reclaimed than would extend the exhaustion date.

Finally, it may be possible to use IP addresses that are currently reserved by IANA. There are proposals to reclaim the class E network addresses;^[10][11] unfortunately, several operating systems and many types of routers would need to be modified or upgraded to make use of these addresses. Many operating systems' TCP/IP stacks, including Microsoft's widely deployed personal computer TCP/IP stack, disallow the use of class E IP addresses, resulting in configuration errors when attempting to assign the address to a host and refusing to communicate with hosts utilizing such an address.^[12][13][14] Similar TCP/IP implementations in many switches and routers also prohibit the use of the class E space.^[15][16] For this reason, the proposal seeks not to redesignate the class E space for public assignment, but instead looks to change the status of the class E range from "Reserved" to "Limited Use for Large Private Internets." This would allow the use of the class E space on large, private networks that require more address space than is currently available through RFC1918.

ISP wide NAT

Similar to how many companies use NAT for most employee computers, an ISP can use NAT for most customers instead of giving them publicly routable dynamically assigned IP addresses. This has many cost-saving and revenue-enhancing advantages to the ISP, including dramatically reducing their need for IPv4 addresses, easier blocking of 'unauthorized' servers running on customer computers such as file sharing systems, the use of web proxies to reduce bandwidth usage and add new banner ads, control of which enhanced services are allowed such as VoIP and games, benefits of customer-wide firewalls, enforcement of laws covering content and tracking, etc. ISPs may allow customers to purchase, at an extra cost, publicly routable dynamic IP addresses similar to how they currently allow, at an extra cost, static IP addresses.

On the other hand, this creates a burden for the ISP to run the NAT services in a law conforming way. Many countries have strict laws on monitoring the users' traffic and behavior (data protection act). Implementing anything more than a pure NAT service, even if it is only with basic traffic logging facility, could put the ISP on the losing end in terms of lawful behavior. A NAT without logging of connections would make it impossible to trace sources of abuse. The ISP does not want to be the Internet Police, nor has it the authority to be that. Its role is the forwarding service of data packets, like the traditional telco does for phone calls. An implementation of ISP wide NAT would require considerable amount of legal consulting to not harm the ISP, which makes this idea difficult to implement.

This idea, however, has been already successfully implemented in some countries like Russia, where virtually all high speed ISPs now have ISP-wide NAT in place, with an option of assigning a publicly routable static IP address at an additional cost.

Markets in IP addresses

The creation of markets to buy and sell IPv4 addresses has been proposed many times as an efficient means of allocation. The primary benefit of an address market would be that IPv4 addresses would continue to be available, although the market price of addresses would be expected to rise over time. These schemes have major drawbacks^{[Neutrality disputed — ]} that have prevented their implementation, as outlined in RFC 2008:

• The creation of a market in IPv4 addresses would only delay the practical exhaustion of the IPv4 address space for a relatively short time, as absolute exhaustion of the IPv4 space would follow within at most a couple of years after the exhaustion of addresses for new allocations.^{[Neutrality disputed — ]}
• The concept of legal "ownership" of IP addresses as property is questionable and it is not even clear which country's legal system lawsuits would be resolved in.
• The administration of such a scheme would be incompatible with current working practices.
• Ad-hoc trading in addresses would lead to fragmented patterns of allocation that would vastly expand the routing table^{[dubious – discuss]}, resulting in severe routing problems for many networks which still use older routers with limited FIB memory or low-powered routing processors. This large cost placed on everyone who uses the internet by those that buy/sell IP addresses is a negative economic externality that any market would need to correct for.
• Trading in IP blocks that are large enough to prevent fragmentation problems would reduce the number of potentially tradeable goods to a few million at most^{[dubious – discuss]}.
• The cost of changing from one set of IP addresses to another is very high, reducing the market liquidity. Organizations that can potentially reorganize their IP addresses usage to free them up so that they can be sold will demand a high price and once bought, will not be resold without a large profit. The cost of renumbering an organization's IP address space each time is comparable to the cost of switching to IPv6 once.
• IP addresses are numbers, so there is no intrinsic value of an IP address. Trading in goods with no intrinsic value (e.g. paper money) instead of goods with extrinsic value (e.g. gold coins) can be risky and requires a stable market.
• Creation of a market requires a critical mass of buyers and sellers. Without that, there will not be price stability. And without an expectation of price stability, it is unlikely that companies will support formation of such a market.

IPv6 as a long-term solution

IPv6 is intended to be the long-term solution to the IPv4 address shortage. Instead of a 32 bit address, with 4.3 billion possibilities, IPv6 represents addresses as 128 bit addresses, providing 3.4×10³⁸ or logically 50 octillion for each of the roughly 6.5 billion people on Earth. However, IPv6 has many problems to address before it is ready.

References

External links

• Official current state of IPv4 /8 allocations, as maintained by IANA
• http://xkcd.com/195/ XKCD web-comic map of the internet, IPv4 address space, 2006.
• IPv6.com - Knowledge Center for Next Generation Internet IPv6