- ¶ 2 Leave a comment on paragraph 2 0
- Monitoring IPv6
- IPv6 deployment status 2016
3. FACTS AND FIGURES – IPV6 DEPLOYMENT
¶ 4 Leave a comment on paragraph 4 0 The global Internet keeps on growing and changing. Estimates predict that by 2020 52% of the world population or 4.1 billion people will be using the Internet. The IP traffic is expected to triple between 2016 and 2020. It is predicted that by 2020 there will be 26.3 billion networked devices and connections globally, 5.5 billion global mobile users and 11.6 billion mobile-ready devices and connections.
¶ 5 Leave a comment on paragraph 5 1 It becomes increasingly important for the Internet to accommodate scale. IPv6 will enable the Internet to cope with the huge demand for IP addresses in the future.
¶ 7 Leave a comment on paragraph 7 0 This chapter makes a status update of the IPv6 deployment. It looks at the IPv6 readiness of the Internet infrastructure and will assess the uptake in IPv6 usage. As a start, section 3.2 is an non-exhaustive overview of indicators, measurements and tools that are being used to monitor progress in IPv6 deployment.
3.2. Monitoring IPv6
¶ 9 Leave a comment on paragraph 9 0 Deploying IPv6 means getting the infrastructure ready and starting to use the IPv6 protocol to communicate over the Internet. Both go hand in hand and obviously, the second can’t happen if the first hasn’t been realised.
¶ 11 Leave a comment on paragraph 11 0 Both aspects of the IPv6 deployment are being monitored. Organisations keep track of the readiness of the infrastructure – the core internet infrastructure as well as the user equipment and applications – and of the amount of IPv6 traffic that is sent over the infrastructure by those capable to do so. This section first looks at different ways to assess the IPv6 readiness and then at different ways in which IPv6 traffic is being measured.
3.2.1. Assessing IPv6 readiness
¶ 12 Leave a comment on paragraph 12 1 IPv6 ready means that it is possible to communicate over IPv6. This requires that the infrastructure, the machines and applications are capable of handling IPv6. IPv6 readiness starts with the sender’s and the receiver’s equipment and software, and includes everything in between. A detailed assessment of the progress in IPv6 deployment needs to take the whole chain into account. Only when it is technically possible to have IPv6 traffic, monitoring IPv6 traffic makes sense. What follows are frequently used indicators of IPv6 readiness.
* The allocation of IPv6 address blocks
¶ 13 Leave a comment on paragraph 13 0 The public Internet is composed of Autonomous Systems (AS) or individual networks that exchange IP traffic, and each network has an unique Autonomous System Number (ASN). Typical examples of an AS are the network operated by an ISP or by a large organisation.
¶ 15 Leave a comment on paragraph 15 0 The first action for a network operator that wants to enable IPv6 on its network – once the operator’s the equipment and infrastructure is IPv6 ready – is to get IPv6 address space. Blocks of IPv6 addresses are allocated by the Regional Internet Registries (RIRs) – via Local Internet Registries (LIRs) – to individual networks (ASes). Assuming that an operator will only request IPv6 addresses when his network is capable of handling IPv6, the demand for IPv6 addresses, measured by the number of IPv6 address blocks and the volume of address space that is allocated by the RIRs serves as an indicator for IPv6 readiness of network operators.
¶ 17 Leave a comment on paragraph 17 0 The RIRs publish statistics on the allocation of IPv6 blocks for their region. The table below gives an overview of the number of IPv6 allocations by each RIR for the last 10 years.
¶ 22 Leave a comment on paragraph 22 0 Network operators can choose between different sizes of IPv6 address blocks with minimum of one /32 address block per allocation by a RIR. The total volume of allocated addresses shows a different dynamic between the regions. The below table shows the volume of allocated IPv6 address space per year in number of /32 blocks. One /32 block represents an address space of 79,228,162,514,264,337,593,543,950,336 IPv6 addresses. Note that one /32 block is larger than the whole IPv4 space. By October 2016, the total volume of IPv6 space given out was 202,660.02 /32 blocks. Although this is more than 202,600 times the IPv4 Internet space, it only represents 0.038% of the available IPv6 space.
|IPv6 Addresses (/32s)||2006||2007||2008||2009||2010||2011||2012||2013||2014||2015|
* Routable IPv6 networks
¶ 26 Leave a comment on paragraph 26 0 The first part of an IPv6 address, the prefix, specifies the network, while the remaining part specifies a particular address in that network. For a network to be reachable over IPv6, this prefix must be visible on the Internet, i.c. a network must announce an IPv6 prefix in the routing table. RIPE NCC is measuring the percentage of IPv6 enabled networks that announce an IPv6 prefix. The data is published in an online graph, adaptable per country or per groups of countries: http://v6asns.ripe.net/v/6.
* IPv6 RIPEness
¶ 27 Leave a comment on paragraph 27 0 IPv6 RIPEness is a tool developed by the RIPE NCC to monitor and assess the IPv6 readiness amongst the RIPE NCC members (LIRs). By marking specific milestones in the deployment process, such as requesting an IPv6 allocation and making the prefix visible on the Internet, the organisation itself as well as other stakeholders can see the high-level progress in IPv6 deployment based on comparable criteria.
* End user IPv6 readiness
¶ 29 Leave a comment on paragraph 29 0 APNIC Labs has designed a test system that reports on end-user capability based on daily tests of random internet users. The APNIC measurements are publicly available. The webpage gives a global overview and statistics on a regional and country by country level: http://stats.labs.apnic.net/ipv6/
* Deployment ratios
¶ 32 Leave a comment on paragraph 32 0 Efforts have been made to develop IPv6 deployment matrices, to define overall IPv6 deployment levels and allow for comparing between countries and regions. Cisco calculates an overall IPv6 deployment ratio based on three other matrices; IPv6-enabled transit AS, IPv6 content and IPv6 users.
3.2.2. Measuring IPv6 traffic
* Global operators and content providers
¶ 34 Leave a comment on paragraph 34 0 Global content providers, service providers and operators that have enabled IPv6 for their networks and services monitor and report on the IPv6 traffic. Google tracks on an ongoing basis the percentage of users that access Google over IPv6. Statistics are published per country: https://www.google.com/intl/en/ipv6/statistics.html
* IPv6 Domain name system
¶ 35 Leave a comment on paragraph 35 0 Each Top Level Domain (TLD), like .com, .org or .de, has its own authoritative nameservers which contain the information on their zone. To support IPv6, these nameservers should have an IP address themselves and native IPv6 connectivity so that they can be reached over IPv6; have AAAA records for their IPv6 address (glue records) in the root zone; and be able to return AAAA (IPv6) address records. Daily statistics on these three requirements are generated by Mike Leber:
¶ 38 Leave a comment on paragraph 38 0 Domain name registries can count the number of domain names in their zone that have an IPv6 address (AAAA-record) and can track the number of DNS queries they receive over IPv6. CZ.NIC, for example, shows these statistics for the .cz domain names on its website: https://stats.nic.cz/stats/ipv6_domains/?rd=2016-09-30&dr=1y&tp=i-1m&ss=0&ds=normal&da=chart
* Internet Exchange Points (IXPs)
¶ 39 Leave a comment on paragraph 39 0 Internet Exchange Points (IXPs) enable the interconnection and exchange of IP-traffic between Autonomous Systems (networks). IXPs can monitor the amount traffic over IPv6 that passes through the IXP. The Amsterdam Internet Exchange (AMS-IX) , for example, has real time IPv6 traffic statistics on its website:
3.2.3. Places to monitor IPv6 Adoption
3.3. IPv6 deployment status 2016
3.3.1. Global uptake: historic evolution
¶ 49 Leave a comment on paragraph 49 0 The past ten years saw a yearly increase of the number of IPv6 allocations by the RIRs. For every year however, the number of IPv6 allocations is much smaller than the number of IPv4 allocations. However, allocated IPv6 blocks are on average much larger and as result that the total volume of allocated IPv6 addresses per year is much higher.
3.3.2. Global uptake: IPv6 slowly entering mainstream in 2016?
¶ 56 Leave a comment on paragraph 56 0 The global IPv6 deployment is on the rise and in many ways 2016 has been a remarkable year that shows how IPv6 is slowly entering mainstream usage. There are different indications for this, and sind mid 2016 a rapid growth in IPv6 supported content could be observed.
¶ 62 Leave a comment on paragraph 62 0 At the beginning of 2016 the percentage of users that accessed Google using IPv6 flirted for the first time in history with the 10% threshold. By June 2016 Google already measured more than 12% users accessing the search engine over IPv6, by October 2016 it reached 15% and the trend continued.
¶ 66 Leave a comment on paragraph 66 0 Akamai, one of the world’s leading content delivery networks (CDN) service providers, saw IPv6 increasingly entering the mainstream in 2016 and reported on major movements in deploying IPv6 by many of the top networks and content providers in the world. Of Akamai’s top five network providers by traffic volume, all but one have IPv6 adoption over 20%. Of the top 25 networks by volume, 14 have IPv6 adoption over 10%, and around a third of the top 100 networks by volume have started rolling out IPv6.
¶ 70 Leave a comment on paragraph 70 0 In November 2016 Cloudflare, announced that almost every site using Cloudflare (more than 4 million in total) was using IPv6. Cloudflare reported significant IPv6 traffic globally where networks had enabled IPv6 to the consumer.
¶ 74 Leave a comment on paragraph 74 0 During the first half of 2016 RIPE NCC counted for the first time ever more than 25% of networks (Autonomous Systems AS) within its service region that announced one or more IPv6 prefixes. This was only 5% in 2009 and 15% at the beginning of 2012.
¶ 78 Leave a comment on paragraph 78 0 By September 2016, one year after full IPv4 depletion for the North American region (24 September 2015), more than half of the networks in the ARIN membership had registered IPv6 addresses. ARIN reported that it continued issuing IPv6 address blocks to 60-100 additional organisations per month.
¶ 86 Leave a comment on paragraph 86 0 AT&T began planning for the transition to IPv6 in 2006. As of September 30, 2016, approximately 60% of the wireline traffic and nearly 15% of the wireless traffic originating from AT&T Autonomous System Numbers (ASNs) was using IPv6.
¶ 90 Leave a comment on paragraph 90 0 In October 2016 a blogpost compared the use of IPv6 in the world of web objects to that of the infrastructure of the DNS, and concluded that the DNS had seen significant progress in the adoption of IPv6, and slightly more than one third of all users was capable of resolving names using IPv6, as compared with a 7% measurement of users capable of using IPv6 in fetching objects over the web. This lead to the conclusion that the DNS was well on the path of transition and perhaps further than other elements of the Internet’s infrastructure.
3.3.3. Regional trends
¶ 92 Leave a comment on paragraph 92 0 This section gives an overview of the current state of IPv6 deployment in the world per geographical region and lists the top 50 countries by IPv6 deployment. The next chapter, will exhibit experiences from countries with a high deployment per region and per sector. It is good practice to only compare data over time and between regions that is based on the same or very similar methodology. The BPF agreed to use the APNIC Labs statistics for this section.
¶ 94 Leave a comment on paragraph 94 0 According to the APNIC Labs measurements for mid November 2016, the global IPv6 capability was close to 8%. The Americas (18% IPv6 capable) and Europe (12% IPv6 capable) scored above the global average. IPv6 capability in the other regions lays below the global average. The country-by-country comparison in this section will show that there are huge differences within the regions and that a few countries are boosting the regional score, for example the IPv6 capability in the USA is almost as high as twice the regional score for the Americas and in only one other country, Peru, the IPv6 capability is (slightly) higher than the average.
|IPv6 Capable||IPv6 Preferred|
¶ 99 Leave a comment on paragraph 99 0 Of the top-50 countries ranked by IPv6 capability 17 showed a double digit deployment rate in October 2016. Of these 17 countries, 10 are European, 3 Latin American, 2 from the Asia Pacific region and 2 from North America North America.
¶ 101 Leave a comment on paragraph 101 0 Only one country, Belgium (56% IPv6 capable), scored higher than 50% on IPv6 capability in October 2016 and with 46% Belgium also leads the ranking in terms of IPv6 use ratio. On some distance, Belgium is followed by the US (34% IPv6 capable – 31% IPv6 use ratio) and Switzerland (31% IPv6 capable – 27% IPv6 use ratio).
¶ 103 Leave a comment on paragraph 103 0 Ecuador, Peru and Brazil are leading in Latin America, with between 10% and 20% IPv6 capability and use ratio. From the Caribbean, Trinidad and Tobago is flirting with the 10%.
¶ 105 Leave a comment on paragraph 105 0 Japan (16% IPv6 capable – 14% use ratio) and Malaysia (15% capable – 15% use ratio) show the highest IPv6 deployment in Asia Pacific. All other countries from the region have deployment rates lower than 10%.
¶ 109 Leave a comment on paragraph 109 0 Early October 2016, no African country scored higher than 1% on IPv6 capability. On 28 September, however, an important provider in Zimbabwe turned on IPv6 with as result that one month later Zimbabwe is leading on the continent with 2.75% IPv6 capable, and 5.28 % IPv6 use ratio.
|Global ranking IPv6 Capable||CC||Country||IPv6 Capable||IPv6 Preferred||IPv6 Use ratio|
|1||BE||Belgium, Western Europe, Europe||55.64%||50.17%||45,89%|
|3||CH||Switzerland, Western Europe, Europe||30.78%||29.27%||26.61%|
|4||GR||Greece, Southern Europe, Europe||27.51%||26.90%||23.00%|
|5||DE||Germany, Western Europe, Europe||27.05%||25.30%||25.20%|
|6||LU||Luxembourg, Western Europe, Europe||23.23%||21.82%||19.41%|
|7||PT||Portugal, Southern Europe, Europe||23.19%||22,39%||19.28%|
|8||GB||United Kingdom, Northern Europe, Europe||20.68%||19.70%||22.16%|
|11||EE||Estonia, Northern Europe, Europe||17.17%||16.74%||15.29%|
|15||FR||France, Western Europe, Europe||13.87%||13.22%||11.51%|
|16||FI||Finland, Northern Europe, Europe||12.63%||11.08%||10.68%|
|19||CZ||Czech Republic, Eastern Europe, Europe||9.41%||8.41%||8.37%|
|20||NO||Norway, Northern Europe, Europe||9.23%||8.33%||7.33%|
|21||IE||Ireland, Northern Europe, Europe||8.40%||8.07%||8.20%|
|23||NL||Netherlands, Western Europe, Europe||7.37%||6.82%||6.75%|
|25||RO||Romania, Eastern Europe, Europe||6.48%||6.23%||5.73%|
|26||AT||Austria, Western Europe, Europe||5.62%||5.30%||6.59%|
|30||HU||Hungary, Eastern Europe, Europe||4.72%||4.59%||4.85%|
|31||SE||Sweden, Northern Europe, Europe||4.65%||4.23%||3.41%|
|32||AX||Aland Islands, Northern Europe, Europe||3.79%||3.75%||5.92%|
|33||BA||Bosnia and Herzegovina, Southern Europe, Europe||2.89%||2.85%||2.53%|
|36||PL||Poland, Eastern Europe, Europe||2.06%||2.00%||1.70%|
|38||RU||Russian Federation, Eastern Europe, Europe||1.84%||1.79%||2.22%|
|41||SI||Slovenia, Southern Europe, Europe||1.61%||1.59%||1.64%|
|47||DK||Denmark, Northern Europe, Europe||0.93%||0.86%||0.94%|
|48||IT||Italy, Southern Europe, Europe||0.83%||0.80%||1.40%|
|49||BG||Bulgaria, Eastern Europe, Europe||0.73%||0.72%||0.74%|
|50||LV||Latvia, Northern Europe, Europe||0.64%||0.64%||0.06%|
|2||US||United States of America, Northern America, Americas||33.76%||31.08%||33.10%|
|14||CA||Canada, Northern America, Americas||14.04%||13.09%||15.41%|
|Latin America & Caribbean|
|9||EC||Ecuador, South America, Americas||19.18%||18.35%||18.91%|
|10||PE||Peru, South America, Americas||18.69%||17.99%||17.85%|
|17||BR||Brazil, South America, Americas||10.24%||9.62%||11.02%|
|18||TT||Trinidad and Tobago, Caribbean, Americas||9.95%||9.62%||11.06%|
|28||BO||Bolivia, South America, Americas||5.27%||4.65%||5.02%|
|12||JP||Japan, Eastern Asia, Asia||15.81%||14.00%||13.78%|
|13||MY||Malaysia, South-Eastern Asia, Asia||14.53%||13.26%||14.00%|
|22||AU||Australia, Australia and New Zealand, Oceania||7.68%||7.13%||7.25%|
|24||SG||Singapore, South-Eastern Asia, Asia||7.00%||6.07%||4.33%|
|27||IN||India, Southern Asia, Asia||5.54%||3.74%||7.54%|
|34||NZ||New Zealand, Australia and New Zealand, Oceania||2.74%||2.59%||2.40%|
|37||LK||Sri Lanka, Southern Asia, Asia||2.01%||1.91%||1.73%|
|39||TW||Taiwan, Eastern Asia, Asia||1.65%||1.32%||2.37%|
|40||HK||Hong Kong Special Administrative Region of China, Eastern Asia, Asia||1.62%||0.19%||1.16%|
|42||TH||Thailand, South-Eastern Asia, Asia||1.32%||1.25%||1.81%|
|43||KR||Republic of Korea, Eastern Asia, Asia||1.21%||0.39%||1.29%|
|46||VN||Vietnam, South-Eastern Asia, Asia||0.93%||0.82%||2.50%|
|49||CN||China, Eastern Asia, Asia||0.35%||0.19%||0.68%|
|29||SA||Saudi Arabia, Western Asia, Asia||4.73%||4.40%||4.20%|
|35||TR||Turkey, Western Asia, Asia||2.13%||0.01%||0.40%|
|44||IL||Israel, Western Asia, Asia||1.14%||0.98%||2.94%|
|**||ZW||Zimbabwe, Western Africa, Africa||2.75 %||2.68 %||5,28%|
|45||LR||Liberia, Western Africa, Africa||0.95%||0.82%||2.26%|
|(-58-)||SD||Sudan, Northern Africa, Africa||0.19%||0.18%||0.14%|
|(-62-)||BW||Botswana, Southern Africa, Africa||0.12%||0.07%||0.01%|
3.1.4. IPv6 deployment versus Economic performance
¶ 114 Leave a comment on paragraph 114 0 The previous section shows that in general the higher adoption rates can be found in more developed regions and an overall low IPv6 deployment in developing nations of the global south. However there are also huge differences within the regions and between countries with a similar development level.
¶ 116 Leave a comment on paragraph 116 0 There are outliers, for example Peru and Ecuador in South America, and large differences between the national deployment rates within the European Union. These differences in IPv6 deployment don’t line up with the size or strength of economy of the ‘IPv6 leaders’.
¶ 118 Leave a comment on paragraph 118 0 It seems unlikely that there is a correlation between a country’s economic performance (based on indicators such as the GDP) and the level of IPv6 deployment. Comparable markets behave differently and even in the same region and markets, comparable operators that use similar technology and equipment will make different choices to IPv6 deployment. A presentation at RIPE 73 meeting in Oct 2016 provides some data of Ipv6 deployment in comparison to GDP: Are We There Yet? IPv6 as Related to GDP per Capita https://ripe73.ripe.net/presentations/101-IPv6-GDP-ripe73.pdf .
¶ 120 Leave a comment on paragraph 120 0 This observation suggests that economic performance is not the main clarification of the more successful IPv6 deployment in some countries. It seems however that IPv6 deployment in some countries mainly depends on individual decisions by one or more large ISP or operators to deploy IPv6 on their network, for example in Europe and the United States, or are the result of collaboration by industry players and/or encouragement by the government to key decision makers in some Asia Pacific countries. These suggestions seem to be confirmed by the observation that in the Spring of 2015 94% of IPv6 users in the world, confined to just 5 of the world’s 30 largest ISPs or the results in for example Japan.
¶ 122 Leave a comment on paragraph 122 0  Local Internet Registry (LIR) are responsible for the distribution of address space and registration of the address space on a local level. LIRs also ensure that policies and procedures are followed on the local level. Organisations that become LIRs are mainly Internet Service Providers (ISPs) that assign and allocate address space onto their customers, telecom and enterprise organisations, as well as academic institutions.
¶ 124 Leave a comment on paragraph 124 0  The size of a block of addresses is specified by writing a slash (/) followed by a number in decimal which value indicates the length of the network prefix in bits. For example, an address block with 48 bits in the prefix is indicated by /48 and contains 2^(128 − 48) = 2^(80) addresses. The smaller the value of the network prefix, the larger the block: a /21 block is 8 times larger than a /24 block.
¶ 127 Leave a comment on paragraph 127 0  There’s a double effect; IPv6 is allocated in larger blocks of which the size is a matter of choice for the requesting party and special policies came into force to allocate the last /8 IPv4 blocks by APNIC, APNIC and RIPE NCC.
¶ 137 Leave a comment on paragraph 137 0 The same article noted that ‘These 30 ISPs together serviced 42% of the entire internet population, and if ‘these 30 providers were to achieve an average 50% IPv6 uptake in their customer base, then the total IPv6 capability level across the entire Internet would be 20% today, rather than 3.6%’