- ipv6-wg - mailman.ripe.net

Ars Technica article about the last 20^H^H10% of IPv4 space
by Shane Kerr 22 Jan '10

22 Jan '10

All, I hate to interrupt our normal process discussions, but I liked this article by Iljitsch van Beijnum on Ars Technica yesterday: http://tinyurl.com/yaqu3f9 Maybe not news for folks here, but a nice explanation of different views of how much space is actually available. Sadly, I think the summary is correct: "it's now too late for a smooth IPv4-to-IPv6 transition, so prepare for a bumpy one where we run out of IPv4 addresses before there is widespread IPv6 adoption." Fasten your seatbelts everyone! -- Shane

1 0

working group co-chair(s)
by David Kessens 22 Jan '10

22 Jan '10

All, During the charter discussion, some of you suggested to appoint a co-chair. Most RIPE working groups have more than one chair person. In the past, I have never felt the necessity to do this job with more than one person. However, in the spirit that it is always good to try something new or different, we can certainly add a co-chair if the working group believes this to be beneficial for achieving our goals. Considering that we don't have any formal process for appointing a working group chair, there is no standard on how this is achieved in a transparent and fair way. This might have been acceptable in the early days of RIPE when there were far fewer people involved and governments weren't watching us. Given that there is an interest in appointing an additional cochair, this seems an excellent opportunity to achieve something more than just appointing a co-chair: if we define an open process and are happy with the results, we can propose the process as a RIPE policy proposal to the wider community for the appointment of chair people within RIPE. At the same time, I don't believe we should overregulate and loose one of the key benefits of our loosely organized community. Therefore, my personal preference would be to formulate a set of principles that we need to follow to appoint chair people while the actual implementation can vary depending on the position that needs to be filled. I would first like to hear what people think about this. If there is support, I would like to form a small task force to write a first version of the principles and how we plan to implement them in the ipv6 working group. David Kessens ---

15 22

New charter approved
by David Kessens 14 Jan '10

14 Jan '10

The Last Call for our new charter resulted in a very large number of positive responses and nobody at all speaking up against the new charter. I therefore conclude that we have consensus on the charter. I did receive a few comments from the working group chair list that the first sentence mentioning the 'IP protocol' is a bit awkward as 'IP' already stands for 'Internet Protocol'. It was suggested that we simply use the words 'Internet Protocol' instead of 'IP protocol' to avoid this issue. As this is really only an editorial change, I will go ahead and make this change when I request the RIPE NCC to post the new charter. Please let me know if you have an issue with this editorial change. I would like to thank Shane as the author of the charter text and everybody else who contributed with comments and suggestions for the charter discussion. David Kessens --- New proposed charter for the IPv6 working group: (first sentence was changed from: IPv6 is the next generation IP protocol to: IPv6 is the next generation Internet Protocol) ----------------------------------------------------------------------- IPv6 is the next generation Internet Protocol. The IPv6 working group exists to promote IPv6 adoption. The working group activities may be anything useful in helping people to deploy IPv6, and to manage IPv4/IPv6 co-existence. These activities include: * Outreach * Education * Sharing deployment experiences * Discussing and fixing operational issues The working group will cooperate with operators and others, both inside and outside the networking industry, to share resources and combine efforts. -----------------------------------------------------------------------

3 2

2009 IPv6 Address Use Report
by Iljitsch van Beijnum 02 Jan '10

02 Jan '10

[ (Non-cross)posted to NANOG, PPML, RIPE IPv6 wg, Dutch IPv6 TF. Web version for the monospace font impaired and with some links: http://www.bgpexpert.com/addrspace-ipv6-2009.php ] 2009 IPv6 Address Use Report Since 2005, I've been compiling an IPv4 address use report every year. With the start of the new decade, this is a good moment to start doing the same thing for IPv6. http://www.bgpexpert.com/addrspace-ipv6.php shows the amount of IPv6 space given out by RIR and by year, while http://www.bgpexpert.com/ipv6addressespercountry.php shows the amount of IPv6 address space by country. Both these pages are updated weekly from the delegation data that the RIRs publish on their FTP servers. 2008 saw a huge increase in the amount of IPv6 space given out and then a big drop in 2009 (amounts of IPv6 address space in the equivalent of /32s): 1999 10.88 /32s in 17 blocks 2000 19.75 /32s in 32 blocks 2001 33.13 /32s in 61 blocks 2002 156.75 /32s in 271 blocks 2003 261.38 /32s in 290 blocks 2004 13340.63 /32s in 295 blocks 2005 26985.00 /32s in 245 blocks 2006 9798.00 /32s in 243 blocks 2007 6687.01 /32s in 491 blocks 2008 81012.02 /32s in 886 blocks 2009 1091.03 /32s in 1280 blocks However, this is not the complete picture. The large number for 2008 is the result of two unusual events. The first one is LACNIC's delegation of the 2804::/16 block to the Brazilian national internet registry (NIR). At some point in the future, the delegation records will not show such blocks as "used" in their entirety anymore. Also, ARIN delegated 14 /22 blocks within the range 2608::/13 to the US Department of Defense. With these two artifacts removed, the amount of IPv6 space given out per year looks like this: 1999 11.00 /32s in 17 blocks 2000 20.00 /32s in 32 blocks 2001 33.00 /32s in 61 blocks 2002 157.00 /32s in 271 blocks 2003 261.00 /32s in 290 blocks 2004 13341.00 /32s in 295 blocks 2005 26985.00 /32s in 245 blocks 2006 9798.00 /32s in 243 blocks 2007 6687.00 /32s in 491 blocks 2008 1140.00 /32s in 871 blocks 2009 1091.00 /32s in 1280 blocks So the number of blocks given out keeps increasing, but their size is going down. There are two reasons for this: roughly between 2004 and 2006, RIPE and APNIC gave out some very large blocks to some very large ISPs. They mostly stopped doing that. And provider independent blocks started to be allowed and are getting more and more popular. These are the /32 - /35 allocations. /32 is the minimum block size given out to ISPs, this used to be /35. So this view shows the numbers of small-to-medium sized ISPs obtaining IPv6 address space: 1999 11.00 /32s in 17 blocks 2000 20.00 /32s in 32 blocks 2001 33.00 /32s in 55 blocks 2002 157.00 /32s in 254 blocks 2003 223.00 /32s in 251 blocks 2004 235.00 /32s in 241 blocks 2005 217.00 /32s in 217 blocks 2006 186.00 /32s in 186 blocks 2007 351.00 /32s in 351 blocks 2008 734.00 /32s in 734 blocks 2009 1011.00 /32s in 1013 blocks These are the blocks larger than /32 that go to large ISPs (excluding the BR NIR and DoD blocks): 2003 38.00 /32s in 3 blocks 2004 13106.00 /32s in 9 blocks 2005 26768.00 /32s in 13 blocks 2006 9612.00 /32s in 14 blocks 2007 6336.00 /32s in 8 blocks 2008 406.00 /32s in 7 blocks 2009 80.00 /32s in 4 blocks And these are the blocks smaller than /35, which are now mostly provider independent blocks, but also "critical infrastructure", such as root servers get a /48 block: 2001 0.00 /32s in 6 blocks 2002 0.00 /32s in 17 blocks 2003 0.00 /32s in 36 blocks 2004 0.00 /32s in 45 blocks 2005 0.00 /32s in 15 blocks 2006 0.00 /32s in 43 blocks 2007 0.00 /32s in 132 blocks 2008 0.00 /32s in 130 blocks 2009 0.00 /32s in 263 blocks So after a small dip in 2006, the number of small-to-medium sized ISPs obtaining IPv6 address space shows a steady upward trend, but apparently the very large ISPs either already got their IPv6 address space, are not focussing on IPv6 right now, are starting with a small block (or several small blocks), or a combination of all of these factors. Even with the BR NIR and DoD blocks included, the (equivalent of) 39395.56 /32s given out in 4111 blocks is only 0.026% of the 536870912 possible /32s in the currently defined global unicast space (2000::/3). For comparison, the number of IPv4 blocks given out is 99562.

1 0

2009 IPv4 Address Use Report
by Iljitsch van Beijnum 01 Jan '10

01 Jan '10

[ (Non-cross)posted to NANOG, PPML, RIPE IPv6 wg, Dutch IPv6 TF. Web version for the monospace font impaired and with some links: http://www.bgpexpert.com/addrspace2009.php ] 2009 IPv4 Address Use Report As of January first, 2010, the number of unused IPv4 addresses is 722.18 million. On January 1, 2009, this was 925.58 million. So in 2009, 203.4 million addresses were used up. This is the first time since the introduction of CIDR in 1993 that the number of addresses used in a year has topped 200 million. With 3706.65 million usable addresses, 80.5% of the available IPv4 addresses are now in some kind of use, up from 75.3% a year ago. So the depletion of the IPv4 address reserves is continuing in much the same way as in previous years: Date Addresses free Used up 2006-01-01 1468.61 M 2007-01-01 1300.65 M 167.96 M 2008-01-01 1122.85 M 177.80 M (with return of 16.78 M to IANA) 2009-01-01 925.58 M 197.27 M 2010-01-01 722.18 M 203.40 M These figures are derived from from the Internet Assigned Numbers Authority's IANA IPv4 Address Space Registry page and the records published on the FTP servers of the five Regional Internet Registries (RIRs): AfriNIC, which gives out address space in Africa, APNIC (Asia-Pacific region), ARIN (North America), LACNIC (Latin American and the Caribbean) and the RIPE NCC (Europe, the former Soviet Union and the Middle East). The IANA list shows the status of all 256 blocks of 16777216 addresses identified by the first 8-bit number in the IPv4 address. http://www.bgpexpert.com/ianaglobalpool.php is a graphical representation of the IANA global pool (updated weekly). The RIR data indicates how much address space the RIRs have delegated to internet service providers (and sometimes end-users). The changes over the course of 2009 are as follows: Delegated Blocks +/- 2009 Addresses Used Available to/status (in millions) AfriNIC 2 33.55 14.89 18.66 APNIC 34 +4 570.42 540.36 30.06 ARIN 31 520.09 486.58 33.51 LACNIC 6 100.66 79.77 20.89 RIPE NCC 30 +4 503.32 450.11 53.21 RIRs subtotal 103 +8 1728.05 1571.71 156.34 LEGACY 92 1543.50 1413.88 129.62 UNALLOCATED 26 -8 436.21 436.21 Totals 221 3707.76 2985.59 722.17 The RIRs requested an unusually small number of /8s from IANA: only eight. As a result, APNIC is well below the nine months working inventory threshold, so it should be getting no less than six additional /8s soon to get back to 18 months working inventory. Similarly, ARIN should be getting five additional /8s soon. This would bring us to 15 /8s remaining in the IANA global pool, and should allow for regular operation for about the rest of the year. Then in early 2011, the next round of delegations will have to happen, which may or may not hit the magic fifth-to-last /8, after which the remaining four will be given to the other four RIRs and then each RIR will run out of IPv4 space at its own pace. The total number of available addresses is slightly higher than the previously mentioned figure at 3707.76 million because the table above includes 172.16.0.0/12 and 192.168.0.0/16, which are set aside for private use. Networks 0.0.0.0/8 and 127.0.0.0/8 aren't usable because of special uses and 10.0.0.0/8 is also set aside for private use. 224 - 239 are multicast addresses, and 240 - 255 is class E, which is "reserved for future use". The 2985 million addresses currently in use aren't very evenly distributed over the countries in the world. The current top 15 is: 2010-01-01 2009-01-01 increase Country 1 - US 1495.13 M 1458.21 M 2.3% United States 2 - CN 232.45 M 181.80 M 27.9% China 3 - JP 177.15 M 151.56 M 16.9% Japan 4 - EU 149.48 M 120.29 M 24.3% Multi-country in Europe 5 (6) DE 86.51 M 81.75 M 5.8% Germany 6 (9) KR 77.77 M 66.82 M 16.4% Korea 7 - CA 76.96 M 74.49 M 3.3% Canada 8 - FR 75.54 M 68.04 M 11.0% France 9 (5) GB 74.18 M 86.31 M -14.1% United Kingdom 10 - AU 39.77 M 36.26 M 9.7% Australia 11 - BR 33.95 M 29.75 M 14.1% Brazil 12 - IT 33.50 M 29.64 M 13.0% Italy 13 (14) RU 28.47 M 23.18 M 22.8% Russia 14 (13) TW 27.10 M 24.01 M 12.9% Taiwan 15 (17) NL 22.84 M 21.67 M 5.4% Netherlands The reduction in address use by the UK is because net 51.0.0.0/8 is now registered as country "EU" rather than "GB". In 2008, the United States was the biggest user of new address space with 50 million new addresses put into use. China was second with 46.50 million. In 2009, they swapped places: China used up 50.65 million addresses and the US a mere 36.92 million. The US now holds 50.1% of the IPv4 address space in use, down from 52.4% last year. The total for the top 15 excluding the US is now 38%, up from 35.8%. The rest of the world gets the remaining 12%, up from 11.8%. The size of address blocks given out was increasing until 2005, and then started decreasing. The table below shows the number of delegations for a certain range of block sizes (equal or higher than the first, lower than the second value). 2003 2004 2005 2006 2007 2008 2009 < 1000 745 1022 1309 1507 1830 1896 1747 1000 - 8000 1009 1516 1891 2265 2839 3235 3185 8000 - 64k 1014 1100 1039 1192 1015 1129 1169 64k - 500k 215 404 309 419 395 410 403 500k - 2M 46 61 60 57 62 82 70 > 2M 6 7 18 17 24 18 21 In millions of addresses: 2003 2004 2005 2006 2007 2008 2009 < 1000 0.25 0.35 0.44 0.51 0.63 0.65 0.59 1000 - 8000 3.45 4.49 5.07 5.83 6.93 7.75 7.55 8000 - 64k 14.00 15.99 15.46 18.01 15.67 17.40 18.01 64k - 500k 25.51 42.01 34.23 50.86 50.83 52.58 50.50 500k - 2M 31.98 44.63 41.63 46.69 45.50 57.41 49.28 > 2M 12.58 20.97 68.62 52.43 67.37 54.00 54.12 Numbers of delegations and their average size: Year Blocks Addresses (M) Average block size 2000 2794 78.35 28043 2001 2824 88.95 31497 2002 2463 68.93 27985 2003 3035 87.77 28921 2004 4110 128.45 31252 2005 4626 165.45 35765 2006 5457 174.32 31945 2007 6165 186.92 30320 2008 6770 189.79 28035 2009 6595 180.06 27302 (The numbers of addresses given out are lower here because ARIN often attributes the delegation of new addresses to a previous year, this view doesn't correct for that.)

1 0

2009-08 Implementation: IPv6 PI Assignments for LIRs
by Nathalie Trenaman 10 Dec '09

10 Dec '09

[Apologies for duplicate emails] Dear Colleagues, We are pleased to announce that RIPE Policy Proposal 2009-08,"IPv6 Provider Independent (PI) Assignments for LIRs", has been implemented. The RIPE NCC is now ready to accept requests for IPv6 address space under the new policy. This proposal can be found at: http://www.ripe.net/ripe/policies/proposals/2009-08.html The updated "IPv6 Address Allocation and Assignment Policy" document is available at: http://www.ripe.net/ripe/docs/ipv6-policies.html The following document has also been updated to reflect this policy implementation: http://ripe.net/ripe/docs/ipv6-pi-support.html Regards, Nathalie Trenaman Registration Services RIPE NCC

1 0

Last Call: New IPv6 working group charter (20091213)
by David Kessens 07 Dec '09

07 Dec '09

All, We have reached the point where it is time to do a Last Call on the latest revision of the new proposed ipv6 working group charter as attached below and posted to the mailing list earlier today. We have gone from determining the generic direction to fine tuning the language for final publication. As Shane mentioned in his previous mail it looks like we have reached the point where further changes are not going to provide for a much improved outcome. Please speak up on or before 20091213 to voice your support for the new charter, or alternatively a statement why we need to consider further changes with suggestions of what these changes should entail. I have already started to keep track of people who stated their support before the official Last Call started, but after Shane's final version was posted, so there is no need to do so again. Just in case we do find very minor issues like a spelling error, I will correct them without issuing a new Last Call. Barring any other negative comments that preclude me from declaring a rough consensus, I will ask the RIPE NCC to post the new charter on the website after the Last Call has passed. I hope this helps, David Kessens --- New proposed charter for the IPv6 working group: ----------------------------------------------------------------------- IPv6 is the next generation IP protocol. The IPv6 working group exists to promote IPv6 adoption. The working group activities may be anything useful in helping people to deploy IPv6, and to manage IPv4/IPv6 co-existence. These activities include: * Outreach * Education * Sharing deployment experiences * Discussing and fixing operational issues The working group will cooperate with operators and others, both inside and outside the networking industry, to share resources and combine efforts. -----------------------------------------------------------------------

8 7

Proposal for new charter: Version 3
by Shane Kerr 02 Dec '09

02 Dec '09

All, I have tried to incorporate additional comments in this version of the charter. I don't plan on producing any more versions - we can debate each word forever, but I think we have reached the point of diminishing returns. If we have consensus, then we can consider ourselves rechartered! Otherwise someone else can work thorough the text again (or start over from scratch). ----------------------------------------------------------------------- IPv6 is the next generation IP protocol. The IPv6 working group exists to promote IPv6 adoption. The working group activities may be anything useful in helping people to deploy IPv6, and to manage IPv4/IPv6 co-existence. These activities include: * Outreach * Education * Sharing deployment experiences * Discussing and fixing operational issues The working group will cooperate with operators and others, both inside and outside the networking industry, to share resources and combine efforts. ----------------------------------------------------------------------- Differences: -IPv6 is the next-generation IP protocol. The IPv6 working group exists +IPv6 is the next generation IP protocol. The IPv6 working group exists De-hyphenating. Or dehyphenating, if you prefer. ;) -* Discussing operational issues +* Discussing and fixing operational issues I don't see any harm in mentioning that we'll try to fix operational issues in addition to discussing them. -The working group will co-operate within the networking industry and -without, to share resources and combine efforts. +The working group will cooperate with operators and others, both +inside and outside the networking industry, to share resources and +combine efforts. More hyphen fixups. Clarifying that the "networking industry" does not just mean operators. Remove the clumsy "within/without" text that apparently only made sense to me. -- Shane

9 8

RE: [address-policy-wg] IPv6 allocations for 6RD
by Remco van Mook 27 Nov '09

27 Nov '09

Michael key words here are 'average' and 'generous'. Perhaps you should try to read first. Same goes for your statement of counting addresses. In cases like these, I think orders of magnitude are more interesting than narrowly defined numbers. 4 million is roughly 0.1% of the IPv4 pool. Since I think the AVERAGE ISP is probably smaller than that (either in customer count or allocated space), giving all of them 32 bits (or the full IPv4 space) PLUS 4-8 bits of network space (not host space!) for their 6rd plan sounds wasteful to me. The 6rd draft has prefix compression for a reason. Sparse number spaces for expansion and automation are fine. Very sparse number spaces you know from the outset will never be filled or usable for other purposes, are not. That is why deploying multiple instances of 6rd in order to benefit from prefix compression to collapse that required number space without losing functionality sounds like the right direction for me. Remco (and if we decide that a /24 is a better standard to hand out to LIRs than a /32, I can predict today that somebody will come up with a very clever reason to say that /16s are probably even better. Repeat until bored - or out of space.) -----Original Message----- From: address-policy-wg-admin(a)ripe.net [mailto:address-policy-wg-admin@ripe.net] On Behalf Of michael.dillon(a)bt.com Sent: vrijdag 27 november 2009 11:04 To: address-policy-wg(a)ripe.net; ipv6-wg(a)ripe.net Subject: RE: [address-policy-wg] IPv6 allocations for 6RD > Fair enough, I'll bite. Given 2^32(or 4 billion) IPv4 > addresses, and say, 4 million IP's allocated to the average > ISP (I'm being generous > here) there's your 0.1%. The rest of the space will go unused > since we're using 32 bits to identify these sparse blocks in > the v4->v6 translation. Not counting customer /56s, 48s, /60s > or whatever. But customers, and /56s are the essential things to count. Again, you just throw in the number 4 million without explaining where it comes from. IPv4 ISPs come in all sizes with one /24 allocation and some with many allocations of sizes ranging from /17 to /12. Counting IP addresses in IPv6 makes no sense. The addressing hierarchy of IPv6 is designed to have large blocks of unused and unusable addresses. This is both to allow for expansion without changing the network architecture, and to allow for automated address assignment functions which rely on large sparse number spaces to avoid collisions. --Michael Dillon This email is from Equinix Europe Limited or one of its associated/subsidiary companies. This email, and any files transmitted with it, contains information which is confidential, may be legally privileged and is solely for the use of the intended recipient. If you have received this email in error, please notify the sender and delete this email immediately. Equinix Europe Limited. Registered Office: Quadrant House, Floor 6, 17 Thomas More Street, Thomas More Square, London E1W 1YW. Registered in England and Wales No. 6293383.

2 1

Proposal for new charter: Version 2
by Shane Kerr 24 Nov '09

24 Nov '09

All, I've tried to combine all of the input and comments and produce a version of the text which hopefully includes all suggestions. Hopefully it is not too long: ----------------------------------------------------------------------- IPv6 is the next-generation IP protocol. The IPv6 working group exists to promote IPv6 adoption. The working group activities may be anything useful in helping people to deploy IPv6, and to manage IPv4/IPv6 co-existence. These activities include: * Outreach * Education * Sharing deployment experiences * Discussing operational issues The working group will co-operate within the networking industry and without, to share resources and combine efforts. ----------------------------------------------------------------------- Changes: * The 2nd paragraph was updated as per Sander's input. * I took David's point and removed "co-operation" as a activity, since as he indicated it is quite vague without the explanation. Instead I added the sentence at the end there. The idea is to make sure that we note that we have a goal of not duplicating work, but rather use the good stuff that other IPv6 and non-IPv6 organizations have done. If this doesn't make sense in the charter, we can of course leave it out. * I added David's activity about sharing deployment experiences - it seems to make sense. * I added a slightly modified activity based on David's last new one. His was "propose solutions for operational issues". I would rather see this as a more general discussion of operational issues - presumably solutions will be proposed where appropriate. If we want to focus on solutions, then we can revert to David's idea about proposing solutions. I think we are very close, and hopefully people are happy with this one. I don't have any objection to further modifications, although I am looking forward to moving past the rechartering and getting to some planning of what to do with the new-found working group super-powers! :) -- Shane

13 14