- ipv6-wg - mailman.ripe.net

Re: [ipv6-wg] Re: 200 customer requirements for IPv6
by Thomas Narten 18 Nov '05

18 Nov '05

Daniel Roesen <dr(a)cluenet.de> writes: > The folks all having their own /32 already allocated to them and > trying to limit independent address space to "service providers" (or > those who manage to pretend being it, which seems to be easy) are > still arguing for "not for them!" paradigm - easy position with > having their own dishes already done. With all due respect, these sorts of comments are not helpful. They imply bad motivations on some, that IMO, are simply false. There are (IMO) valid reasons for not giving out /32s to everyone. The original (and still compelling, IMO) rational behind the current policy is to ensure good aggregation for the long term, in order to keep the number of distinct prefixes in the DFZ manageable. Hence, the current policy is oriented not towards _all_ ISPs, but those who will (hopefully) have _many_ customers that they assign addresses too. That is, they aggregate addresses for _many_ customers. That is where the "plan for 200 customers" wording originally came from. The presumption is that an ISP/LIR that realistically expects to have 200 customers will be aggregating _many_ customer addresses. If an ISP will have (say) only 100 customers long term, its not at all clear that they will be aggregating significant amounts of address space. Note, I'm not defending the the "200 customer" rule per se; I understand that it is viewed as problematic. But I think the motivation that led to it is still valid, and when folk say "get rid of this requirement," I object to "forgetting" about the original concern when coming up with a replacement policy. > Until we have a clear full replacement (that means a solution which does > NOT ignore real requirements like shim6 does) there should be a very > simple PI policy which issues a /48 or whatever to end sites at a > nominal small fee, paid directly to the RIR. An initial setup fee and a > yearly renewal fee, paid by credit card or something equally simple. > No payment => assignment withdrawn. The initial fee covering the cost of > evaluation of the request, doing the assignment and setting up the > billing account and DNS reverse. The yearly fee covering the maintenance > of the entries in the database and DNS rev NS RRset and the yearly > recurring fee invoicing/billing. Done. Note that ARIN has been and continues to have discussions about how to give out PI space to end sites, but in a way that doesn't explode the routing tables going forward. I think there is (very strong) support for the idea of giving out more PI space, the difficulty has been in coming up with details that people are comfortable won't explode the routing tables going forward. I.e., that in five years we look back and say "oops, we should not have done that, now what do we do" > Setting aside my disbelieve in the FUD about the scalability problem of > real BGP multihoming (noone yet has shown that the relative amount of > multihomers does or will explode), Um, do the math. How many end sites do you think will want to multihome ten years from now? Can you honestly/realistically say it won't be a million? Can the routing infrastructure handle that? (Enough people say "not sure" that IMO it would be foolish to just assume we can do this). Thomas

3 2

Draft Minutes (v1) of IPv6 WG session RIPE51
by David Kessens 22 Oct '05

22 Oct '05

Please see enclosed the first version of the minutes of our session at RIPE51. I plan to declare the minutes final if I don't receive any comments for significant changes by November 4th 2005. I would like to thank Timothy Lowe from the RIPE NCC for taking the minutes. David Kessens --- Minutes IPv6 Working Group session Amsterdam, 12 Oct, 2005 Chair: David Kessens Scribe: Timothy Lowe # of attendees: 95 ACTION 51.1 - Ask the RIPE NCC to revise the acronym list in the registration package to remove obsolete terms as TLA & NLA (David Kessens) Status: Completed ACTION 51.2 - Create a link from the IPv6 WG page to the APNIC events page: http://apnic.net/info/calendar/index.html (RIPE NCC) Status: To be done ACTION 51.3 - Provide traffic reports on ipv6 usage during the RIPE meetings. If available, data from earlier meetings will be made available as well. (RIPE NCC) A. Administrative stuff Appointment of scribe: Timothy Lowe Agenda bashing: No additions to agenda Action point 51.1 was decided. B. Quick update from the RIPE NCC regarding ipv6 services (RIPE NCC) No news from the RIPE NCC C. IPv6 events calendar on the ipv6 working group page (proposed by Timothy Lowe) It was decided that similar event pages already exist and that we can simply provide a link to the APNIC event pages. (See action point D. Discussion on: Global IPv6 routing table status (Gert Doering, Spacenet) Discussion on filtering: Gert Doering: The RIPE Whois database is the only one allowing route6 object creation. Iljitsch Van Beijnum: What is the gain of filtering less specific ? Geoff Huston: What problem are you trying to solve ? Gert Doering: Trying to prevent accident, less maintenance if the documentation is in the database then it will be more stable. Conclusion: as we cannot force people to filter, we will see more specifics anyways (David Kessens) E. Report(s) about *actual* v6 traffic volume as compared to v4? *what's real* out there, not what's on powerpoint? - http://www.sixxs.net/misc/traffic/ (Jeroen Massar) Pointer to Jeroen's website was provided, but content was not discussed. Pim van Pelt offered to do some work to provide more data on his Sixxs ipv6 service but it was decided that the amount of work was probably not worth the effort. Discussed whether any ipv6 usage data from the RIPE meetings itself was available. This discussion resulted in a request and an action point for the RIPE NCC to provide us such data in the future (Action point 51.3) F. The end of the 6bone address space is in sight (Jeroen Massar) Discussion: - what are people going to do with 6bone address space as - per 6/6/6 ? - filter it? - keep it going? - warn people? - what to do with the 6bone registry ? Iljisch v B.: DNS is not auto-configurable in IPv6 like it is in IPv4 so I would like to use 6bone to get an ipv6 assignment for an experiment for well-known ipv6 DNS resolver addresses. Geoff H. : IANA has released a statment that 6bone dies on 06/06/2006. Gerd D.: exactly, what will people do when this happens? No uniform policy on what to do with 6bone prefixes. Some people will start filtering while others won't bother. In general, people advised David Kessens who maintains the 6bone registry to stop accepting registry changes by the close date but to keep the data online for some time to come with proper warning message when information is accessed. G. Developments/initiatives regarding IPv6 in the RIPE region and beyond - 6DISS (Bernard Tuy) Presentation will be available on the RIPE website as soon as Bernard Tuy provides the slides No questions. H. Input for the RIPE NCC Activity Plan (no input from the audience) Z. AOB Peter Koch drew attention to a presentation in the DNS working group regarding the deprecation of IP6.int. ----

2 1

Re: [6bone] RIPE and IPv6 ASN
by marko ramljak 07 Oct '05

07 Oct '05

On Tue, 15 Oct 2002, Arien Vijn wrote: > On 15-10-2002 17:00PM, "Nicolas DEFFAYET" <nicolas.deffayet@localhost > wrote: > > > On Tue, 2002-10-15 at 16:33, Bill Manning wrote: > >> % My company created the first french > >> % IPv6 Internet Exchange, FNIX6 (http://www.fnix6.net) but my company > >> % can't peer on FNIX6 becase we don't have public ASN. > >> > >> this is not the first v6 enabled exchange in France. > >> there are three other exchanges which are v6 enabled. > >> > > > > This other Internet Exchange in France are IPv4 and IPv6 Internet > > Exchange (dualstack). > > IPv6 Internet Exchange = IPv6 only > > Why? Is what we are doing not IPv6? > > Arien IPv6 only = Purely IPv6 (no IPv4 legacy stuff!) I've understood it as a marketing issue... :-) Nicolas, about the problem: If you get LIR Status you'll probably get your ASN quickly. Regards, ./Carlos "Networking is fun!"

2 1

Draft Agenda (v2) for ipv6 wg RIPE51
by David Kessens 05 Oct '05

05 Oct '05

Please see below for the draft agenda (v2) of the ipv6 working group. We still have some room in the agenda so let me know if there is anything that you would like to discuss. As mentioned before, I would also like to draw your attention to the plenary part of the meeting that has various interesting ipv6 presentations on the agenda (see below). David Kessens --- Draft Agenda (v2) for the IPv6 Working Group Meeting RIPE51 When: Wednesday, October 12, 2005, 17:00-18:00 Where: Grand Ballroom, Hotel Krasnapolsky, Amsterdam A. Administrative stuff - appointment of scribe - agenda bashing (David Kessens) B. Quick update from the RIPE NCC regarding ipv6 services (RIPE NCC) C. IPv6 events calendar on the ipv6 working group page (proposed by Timothy Lowe) D. Discussion on: Global IPv6 routing table status (Gert Doering) routing policy: - is it mandatory to register route6's and inet6num's or is it just 'being nice' ? - is the trend going to be that >/32's will be announced as separate /32's? what is the correct plan for shorter than /32 prefixes ? E. Report(s) about *actual* v6 traffic volume as compared to v4? *what's real* out there, not what's on powerpoint? - http://www.sixxs.net/misc/traffic/ (Jeroen Massar) - ... (input from the audience) F. The end of the 6bone address space is in sight (Jeroen Massar) Discussion: - what are people going to do with 6bone address space as per 6/6/6 ? - filter it? - keep it going? - warn people? - what to do with the 6bone registry ? G. Developments/initiatives regarding IPv6 in the RIPE region and beyond - 6DISS (Bernard Tuy) - ... (input from the audience) H. Input for the RIPE NCC Activity Plan (input from the audience) Z. AOB --- Interesting topics in the plenary section: TUESDAY 09:00 - 10:30 6. IPv6 Routing Update. Gert Doering. Spacenet (20 min) Update on observations on the state of the IPv6 default free routing table. 7. IPv6 Multihoming Status. Kurtis Lindqvist. Netnod (30 min) An up-to-date status report on the progress towards scalable IPv6 multihoming. 8. IPv6 Address Allocation --An Alternative Algorithm for the Sparse Allocation Process. Mae Wang (30 min) IP address allocation policies significantly impact the Internet infrastructure, affecting many parties such as router manufacturers, ISPs, and end users. An address allocation policy can also directly affect the performance of the Internet. For example, address fragmentation, a key problem in IPv4, degrades address lookup performance in routers. Thus, a well-designed address allocation policy needs to minimise fragmentation while using the address space efficiently. This paper attempts to quantify the performance of address allocation policies by modelling key features that lead to fragmentation and inefficient address space usage. Our main contributions are: (i) we identify a drawback of the current IPv6 address allocation policy, which treats all entities uniformly, (ii) we propose a scheme that takes future growth rate into account for allocations, and (iii) an analytical model for measuring the efficiency of allocation schemes, allowing us to quantify the improvement our proposal offers over the current scheme. We believe that a quantitative study of allocation policies is timely since IPv6 address allocation is just beginning in earnest.

1 0

RE: [ipv6-wg] Draft Agenda (v1) for ipv6 wg RIPE51
by Mattias.Lignell＠teliasonera.com 03 Oct '05

03 Oct '05

Hello again, I will have a second look at this and discuss the issue internally. I sure hope we can remove the /32's. /// Mattias > -----Original Message----- > From: Jeroen Massar [mailto:jeroen@unfix.org] > Sent: Monday, October 03, 2005 7:15 PM > To: Lignell, Mattias E. > Cc: gert(a)space.net; david.kessens(a)nokia.com; eva(a)telia.net; > Lind, Mikael E.; amar(a)telia.net; ipv6-wg(a)ripe.net > Subject: RE: [ipv6-wg] Draft Agenda (v1) for ipv6 wg RIPE51 > > On Mon, 2005-10-03 at 18:59 +0200, Mattias.Lignell(a)teliasonera.com > wrote: > > Hello all, > > > > About the /32's; they are announced on purpose. The reason > is that we > > have had problems with some ISP's BGP filters that are set to a > > maximum size of /32. As a result did not our IPv6 customers get > > connectivity via ISP's using the /32 size filters. Maybe this is a > > problem only we run in to being one of the first using a very large > > prefix. However, the BGP filters may be updated by now so > we can stop announcing the /32's. > > Thanks for the response, the prefix filtering is not too good > to hear and has been seen before; I know that Daniel Roesen > has been very busy trying to get C&W's /20 available to ISP's > using, amongst others, the output from: > http://www.sixxs.net/tools/grh/compare/?a=2001:5000::/21&b=200 > 1:650::/32 > At the moment apparently Cisco doesn't see the /21... > > This can be easily applied to Telia's /20: > http://www.sixxs.net/tools/grh/compare/?a=2001:2000::/20&b=200 > 1:2040::/32 > > This compares 2001:2000::/20 with 2001:2040::/32. > Taking a quick look it looks quite good, ignoring the light > orange which is there because of the added ASN. There are > only few ISP's who report different paths: > AS 6435 - LavaNet (us / Hawaii) > AS 9264 - Academic Services Network (tw) > AS21155 - Proserve (nl) > AS28788 - Unilogic (nl) {using same upstreams as AS21155) > > You might want to contact these ISP's to ask them + upstreams > to fix their filtering, clearly something is being filtered, > otherwise the paths where the same. > > Greets, > Jeroen > >

1 0

RE: [ipv6-wg] Draft Agenda (v1) for ipv6 wg RIPE51
by Mattias.Lignell＠teliasonera.com 03 Oct '05

03 Oct '05

Hello all, About the /32's; they are announced on purpose. The reason is that we have had problems with some ISP's BGP filters that are set to a maximum size of /32. As a result did not our IPv6 customers get connectivity via ISP's using the /32 size filters. Maybe this is a problem only we run in to being one of the first using a very large prefix. However, the BGP filters may be updated by now so we can stop announcing the /32's. Please note that this has been a temporary work around to maintain connectivity for our customers. We also have old prefixes that we are about to return to RIPE (2001:06C0::/35), but not yet had time to shout down (affects many customers and links). It is of-cause the policy of the TeliaSonera group to normally only announce one prefix. Any comments to this are most welcome. Best regards /// Mattias Lignell -------------------------------------------------------------- Mattias Lignell TeliaSonera Sweden, IP-Networks Phone: +46 8 713 3706 Mobile: +46 70 595 9068 Fax: +46 70 615 6245 Email: mattias.lignell(a)teliasonera.com Address: Vitsandsgatan 9, House-D, 123 86 Farsta, SWEDEN ICQ UIN: 18690075 -------------------------------------------------------------- > ------------------------------------------------------------- > ----- Original Message ----- > From: "Jeroen Massar" <jeroen(a)unfix.org> > To: "Gert Doering" <gert(a)space.net>; "David Kessens" > <david.kessens(a)nokia.com> > Cc: "Amar Andersson" <amar(a)telia.net>; <ipv6-wg(a)ripe.net> > Sent: Friday, September 30, 2005 3:06 PM > Subject: Re: [ipv6-wg] Draft Agenda (v1) for ipv6 wg RIPE51 > > On Thu, 2005-09-29 at 20:46 -0700, David Kessens wrote: One discussion point you might want to raise: - what are people going to do with 6bone address space as per 6/6/6 ? - filter it? - keep it going? - warn people? In january I will send out a note to all the folks still announcing 6bone prefixes to remind them that they should be shutting it down per 6/6/6. There doesn't seem to be a large movement in doing so yet... > C. Discussion on: > Global IPv6 routing table status > (Gert Doering) Maybe something related might be noted here, in the global routing tables I've seen for instance, just a 'random' targetted pick. http://www.sixxs.net/tools/grh/lg/?find=2001:2000::/20 [Notez bien: the question here is about what is going to become the common policy and what is going to happen to the ipv6 routing, it is not to 'attack' or critize an ISP on what they are doing] 2001:2000::/20 announced by AS1299 (TELIANET) 2001:2040::/32 AS3301 (TELIA-SWEDEN) 2001:2060::/32 AS1759 (SONERA-TRANSIT-AS) No route6 for these 3 prefixes and no inet6nums exist for the the two /32's. Which leads me to a couple of questions: - is it mandatory to register route6's and inet6num's or is it just 'being nice' - is the trend going to be that >/32's will be announced as separate /32's? This of course means that the ISP can be found under one block but still affects the number of routes in the routing table. One could filter the smaller ones if the parent route is present. The idea of minimizing routing table size (yes that is not an issue now, but behold the future) is totally gone when ISP's will do this. I do understand why: to make the branches independant and attract the traffic as local as possible. On the otherside there are also ISP's simply requesting multiple separate /32's from the various RIR's. eg UUNet: 2001:0600::/32 - Europe 2001:4440::/32 - Hongkong 2001:4441::/32 - Australia 2001:4442::/32 - Japan 2001:0588::/32 - South Africa (I am ignoring the 6bone block they still have, with 1 other already returned, as that one will disappear anyway) The question in case of this setup is, why didn't they get 1 block? The generic question: what is the correct plan for these setups? > D. Report(s) about *actual* v6 traffic volume as compared to v4? > *what's real* out there, not what's on powerpoint? > (input from the audience) http://www.sixxs.net/misc/traffic/ :) Not much, but it is a bit and for the moment declining a bit weirdly enough. <SNIP> > For example, address fragmentation, a key problem in IPv4, degrades > address lookup performance in routers. Thus, a well-designed > address allocation policy needs to minimise fragmentation while > using the address space efficiently. See my related note above. Greets, Jeroen

2 1

Draft Agenda (v1) for ipv6 wg RIPE51
by David Kessens 30 Sep '05

30 Sep '05

Hi, Please see below for the draft agenda (v1) of the ipv6 working group. This is the first version of the agenda so we are still looking for agenda topics and contributions for the various standard sections of the agenda. Please let me know if there is anything that you would like to discuss. I would also like to draw your attention to the plenary part of the meeting that has various interesting ipv6 presentations on the agenda (see below). David Kessens --- Draft Agenda (v1) for the IPv6 Working Group Meeting RIPE51 When: Wednesday, October 12, 2005, 17:00-18:00 Where: Grand Ballroom, Hotel Krasnapolsky, Amsterdam A. Administrative stuff - appointment of scribe - agenda bashing (David Kessens) B. Quick update from the RIPE NCC regarding ipv6 services (RIPE NCC) C. Discussion on: Global IPv6 routing table status (Gert Doering) D. Report(s) about *actual* v6 traffic volume as compared to v4? *what's real* out there, not what's on powerpoint? (input from the audience) E. Developments/initiatives regarding IPv6 in the RIPE region and beyond (input from the audience) F. Input for the RIPE NCC Activity Plan (input from the audience) Z. AOB --- Interesting topics in the plenary section: TUESDAY 09:00 - 10:30 6. IPv6 Routing Update. Gert Doering. Spacenet (20 min) Update on observations on the state of the IPv6 default free routing table. 7. IPv6 Multihoming Status. Kurtis Lindqvist. Netnod (30 min) An up-to-date status report on the progress towards scalable IPv6 multihoming. 8. IPv6 Address Allocation --An Alternative Algorithm for the Sparse Allocation Process. Mae Wang (30 min) IP address allocation policies significantly impact the Internet infrastructure, affecting many parties such as router manufacturers, ISPs, and end users. An address allocation policy can also directly affect the performance of the Internet. For example, address fragmentation, a key problem in IPv4, degrades address lookup performance in routers. Thus, a well-designed address allocation policy needs to minimise fragmentation while using the address space efficiently. This paper attempts to quantify the performance of address allocation policies by modelling key features that lead to fragmentation and inefficient address space usage. Our main contributions are: (i) we identify a drawback of the current IPv6 address allocation policy, which treats all entities uniformly, (ii) we propose a scheme that takes future growth rate into account for allocations, and (iii) an analytical model for measuring the efficiency of allocation schemes, allowing us to quantify the improvement our proposal offers over the current scheme. We believe that a quantitative study of allocation policies is timely since IPv6 address allocation is just beginning in earnest. ------

3 2

RIPE Policy Development Process
by Paul Rendek 01 Sep '05

01 Sep '05

Dear Colleagues We are pleased to announce that Rob Blokzijl, RIPE Chair, has published a new RIPE Document, "Policy Development Process in RIPE", (ripe-350). This document formally describes the policy development process. The updated process is operational from today, 1 September 2005. A new RIPE web page describes all the active policy proposals and their status. Each proposal has a page showing its history, status and all supporting documentation. This page can be found at: http://www.ripe.net/ripe/policies/proposals/index.html A new mailing list <policy-announce(a)ripe.net> has been created for announcing policy proposals and tracking them through the policy development process. This is not a discussion list; discussions will continue to take place on relevant RIPE Working Group lists and at RIPE Meetings. You can subscribe to the policy-announce list at: http://www.ripe.net/mailman/listinfo/policy-announce List archives are available at: http://www.ripe.net/ripe/maillists/archives/policy-announce/ Kind regards Paul Rendek Head of Member Services and Communications RIPE NCC

1 0

New Document available: RIPE-349
by RIPE NCC Document Announcement Service 27 Jul '05

27 Jul '05

New RIPE Document Announcement -------------------------------------- A new document is available from the RIPE Document store. Ref: ripe-349 Title: Address Space Managed by the RIPE NCC Author: Leo Vegoda RIPE NCC Date: 27 July 2005 Format: PDF=31257 TXT=2986 Obsoletes: ripe-348 Short content description ------------------------- This document details the address space managed by the RIPE NCC and the longest prefixes allocated or assigned from different address ranges. The new version was needed to correct errors in the previous document. Accessing the RIPE Document store --------------------------------- You can access this RIPE Documents in HTML format at: http://www.ripe.net/docs/ripe-349.html The RIPE Document Store is also available via anonymous FTP to ftp.ripe.net, in the directory ripe/docs. The URLs for the new document on the FTP-server are: ftp://ftp.ripe.net/ripe/docs/ripe-349.pdf PDF version ftp://ftp.ripe.net/ripe/docs/ripe-349.txt plain text version

1 0

Re: raw ipv6 sockets - sin6_scope_id field
by Aviraj Saha 25 Jul '05

25 Jul '05

Hi This value needs to be set to the interface index which is returned by the "ip link" command if you are using link local address, and if the 3 rd and 4 th byte is zero. Otherwise the value which is set in the third and fourth byte of the in6_addr structure. Hope this answers your question. With Regards Aviraj Saha

1 0

[ipv6-wg@ripe.net] New IPv6 Address Block Allocated to RIPE NCC
by leo vegoda 22 Jul '05

22 Jul '05

Dear Colleagues, The RIPE NCC received the IPv6 address range 2A01:0000::/23 from the IANA in July 2005. We will begin making allocations from this block in the near future. You may wish to adjust any filters you have in place accordingly. More information on the IP space administered by the RIPE NCC can be found on our web site at: https://www.ripe.net/ripe/docs/ripe-ncc-managed-address-space.html This document has been updated to to provide extra information about the IPv6 address space we manage, following some requests we received. I hope people find the new format useful. Regards, -- leo vegoda Registration Services Manager RIPE NCC

2 3

[ipv6-wg@ripe.net] Allocation behaviour (organisations getting multiple /32's)]
by John Tran 19 Jul '05

19 Jul '05

Hi Jeroen, It is great to know that you are paying attention to allocations that are being made in other region. I would like to clarify the reason why APNIC made muliple allocations to UUNET. UUNET have multiple memberships with APNIC based on the networks in the region. As each economy, membership, managing their own address space therefore they also requesting for IP address separately. For aggregation purpose APNIC normally make allocations from contiguous block, as you can see from these allocations. This is similar to making a large allocation to an organisation then they further splitting between network for each economy. I hope the above explanation answered some of your queries if you need further information please feel free to contact us. Regards Son APNIC > > -------- Original Message -------- > Subject: [ipv6-wg(a)ripe.net] Allocation behaviour (organisations getting > multiple /32's) > Date: Sat, 09 Jul 2005 12:52:33 +0200 > From: Jeroen Massar <jeroen(a)unfix.org> > Organization: Unfix > To: ipv6-wg(a)ripe.net > > I just noticed that UUNET/MCI got an additional 3 /32's: > > inet6num: 2001:4441::/32 > netname: UUNET-AU-NETBLOCK-20050708 > descr: UUNET Australia Limited > descr: UUNET Network > country: AU > > inet6num: 2001:4440::/32 > netname: UUNET-HK-NETBLOCK-20050708 > descr: UUNET Worldcom Hong Kong Ltd > descr: UUNET Network > country: HK > > inet6num: 2001:4442::/32 > netname: UUNET-JP-NETBLOCK-20050708 > descr: UUNET Japan, Ltd > descr: UUNET Network > country: JP > > They where all assigned to 1 person, who apparently works for UUNET/MCI > in the US, not even the APNIC region, not even a role. > > Did somebody say goodbye to aggregation? > > Btw UUNET also has: > 2001:600::/32 (Europe) > > Apparently nothing in the US, in total 4x /32. 4 slots gone away. > Are all global companies going to request separate /32's ? > > Greets, > Jeroen > > > > -- > > _____________________________________________________________________ > Anne Lord, Communications Director <anne(a)apnic.net> > Asia Pacific Network Information Centre (APNIC) Tel: +61-7-3858-3100 > PO Box 2131 Milton, QLD 4064 Australia Fax: +61-7-3858-3199 > --------------------------------------------------------------------- >

3 3

[ipv6-wg@ripe.net] ipv6-enabled webserver
by Thu Thuy 18 Jul '05

18 Jul '05

Hi, all I'm in my way to investigate on ipv6 trial and implementation. Now i'm stucking with a quite big question on ipv6-enabled software. As i saw, ipv6 web services is entirely implemented in almost such large ipv6 research projects as KAME, WIDE, GEANT... In Japan, some ISPs offer ipv6 web hosting to their customers. Many websites support native ipv6 such as RIPE, APNIC... Could you pls show me what kind of webserver software be used in above-mentioned projects and organisations to fully implement ipv6 web service to users. Thanks.

6 9

[ipv6-wg@ripe.net] Allocation behaviour (organisations getting multiple /32's)
by Jeroen Massar 09 Jul '05

09 Jul '05

I just noticed that UUNET/MCI got an additional 3 /32's: inet6num: 2001:4441::/32 netname: UUNET-AU-NETBLOCK-20050708 descr: UUNET Australia Limited descr: UUNET Network country: AU inet6num: 2001:4440::/32 netname: UUNET-HK-NETBLOCK-20050708 descr: UUNET Worldcom Hong Kong Ltd descr: UUNET Network country: HK inet6num: 2001:4442::/32 netname: UUNET-JP-NETBLOCK-20050708 descr: UUNET Japan, Ltd descr: UUNET Network country: JP They where all assigned to 1 person, who apparently works for UUNET/MCI in the US, not even the APNIC region, not even a role. Did somebody say goodbye to aggregation? Btw UUNET also has: 2001:600::/32 (Europe) Apparently nothing in the US, in total 4x /32. 4 slots gone away. Are all global companies going to request separate /32's ? Greets, Jeroen

1 0

[ipv6-wg@ripe.net] Minutes IPv6 WG RIPE 50
by David Kessens 18 Jun '05

18 Jun '05

I did not receive any comments on the draft minutes (v1) by June 10. Therefore, the draft minutes (v1) are now the official minutes of our session during RIPE 50 except for one typo that I found myself, in the last sentence, please replace 'exepected' by 'expected'. Thanks, David Kessens --- ----- Forwarded message from David Kessens <david.kessens(a)nokia.com> ----- Date: Tue, 31 May 2005 20:42:40 -0700 From: David Kessens <david.kessens(a)nokia.com> To: ipv6-wg(a)ripe.net Cc: meeting(a)ripe.net Please see below for our first cut of the minutes for our working group session during RIPE 50. Please let me know if you have any comments or corrections. The minutes will be declared final by Friday, June 10, 2005 if no significant changes (other than typo fixes and minor editorial issues) are found. David Kessens --- Draft Minutes (v1) IPv6 WG RIPE 50 When: Friday, 6 May 2005, 09:00 - 10:30 Where: C58, Clarion Hotel, Stockholm, SE Minute Taker: Adrian Bedford Quick Update from the RIPE NCC Regarding IPv6 Services (Ruud de Kooter - RIPE NCC) After the presentation, Ruud was asked if there were plans to integrate the IPv6 whois interface with the main whois. Currently the proxy cannot talk to the interface and this hampers some functionality (such as rate limiting). Ruud asked Shane Kerr to answer this. Shane explained that the RIPE NCC has a v6 proxy running over a v4 database backend. The RIPE NCC has been looking at how people use the database for two years and patterns suggest that now would be a good time to integrate native IPv6 into the server using /48 as a client identifier. Discussion on: Global IPv6 Routing Table Status (Gert Doering, Spacenet) Gert indicated that the Routing Working Group was discussing best practices for BGP filters. The overall thrust of the presentation was that v6 is growing smoothly with no big disasters so far. A question was asked about ratio comparison between the numbers of Autonomous Systems (AS) announced by IPv6 networks as opposed to v4. The current data suggests that there are 1.419 announced entries per originating AS for v6 and 10.308 for v4. A point was made that Deutsche Telecom (DT), who have one of the largest v6 allocations made to date would not announce their block as one AS, they would be more likely to split this and allocate blocks to different service providers for political reasons. Someone from DT commented that the company has the right to do this, though denied that the move was in any way politically motivated. Renumbering IPv6 Networks (Stig Venaas, Uninett & Tim Chown, University of Southampton) After the presentation, it was mentioned that we are assuming that customers of service providers will keep their space, yet ten years ago major providers did force renumbering onto customers. It was agreed that renumbering of IPv6 is, at least in principal, no different to what happened with v4. Developments/Initiatives Regarding IPv6 in the RIPE Region and Beyond - Francis Dupont: Point6 (http://www.point6.net/ - Gert Doering: A new maillist for discussion of ipv6 operational issues was started: http://lists.cluenet.de/mailman/listinfo/ipv6-ops/ After the announcement of the list, it was agreed that having a publicly available archive of this new list would be a good idea. Although it exists on the site that hosts the list, this is a personal website and for stability purposes, maybe a copy of the archive should be placed elsewhere. It was felt that the RIPE website was unsuitable for this. - Carlos Friacas: o IPv6 - Advanced Network Developments's first year in Portugal o 9K IPv6 Schools' Network Initiative o About the IPv6 Steering Committee Revisiting the /48 recommendation There was a lengthy discussion around RFC 3177. It was made clear that the discussion was not intended to reach final consensus on all the issues involved, but merely to serve as a way to get an idea about the general direction on where the community would like to go to make it possible to write up an initial proposal. Many agreed that the /48 is too much for many users for now and in the future. A discussion followed about how we should define a site with the suggestion that we need to accept that this can remain a grey area, which is not at all well defined right now. Discussion is needed around this topic, though the group tended to accept that it might not be easy to put down limits in black and white. Most people in the room felt that RFC3177 needs a revision. Most people felt that the /48 limit itself doesn't need to be changed, but most agreed that there is a need for a new category that falls somewhere between /48 and /64 for users that have a need for subnetting but for which a /48 seems excessive. This category would fit the mass consumermarket or (very) small office market, however, it was noted that we should base the category on technical needs of the user, not in economic terms like mass consumer market. It is exepected that an Internet Draft will be written that can be discussed at RIPE 51. ----- End forwarded message -----

1 0

[ipv6-wg@ripe.net] documenting RIPE critical infrastructure/micro allocations
by Bjoern A. Zeeb 02 Jun '05

02 Jun '05

Hi, while updating some prefix filters I started a link collection documenting the entries. I have found that other RIRs (ARIN, APNIC) have neat documentation about the critical infrastructure/micro allocations assigned but I only found two references of the Block RIPE uses for v6 IXP assignments. | ARIN Critical infrastructure, micro allocations | http://www.arin.net/reference/micro_allocations.html | APNIC Critical infrastructure, micro allocations | http://www.apnic.net/db/ranges.html For LACNIC I haven't found any assignments but following documentation: | LACNIC Critical infrastructure, micro allocations | http://lacnic.net/en/politicas/ipv6.html 5.5. IPv6 micro allocations | http://lacnic.net/documentos/politicas/micro-asignaciones_ipv6-sp.pdf For RIPE Gert's page is talking about 2001:7F8::/29 and the DB entry. | http://www.ripe.net/whois?searchtext=EU-ZZ-2001-07F8 | http://www.ripe.net/ripe/docs/ipv6-policy-ixp.html Could someone add documentation/web page stating that RIPE is doing /48 assignments from this Block for critical infrastructure (read: IXPs)? Perhaps also giving a list of which prefixes got assigned already as these won't show up in delegated-ripencc-latest. Greetings Bjoern A. Zeeb -- Bjoern A. Zeeb bzeeb at Zabbadoz dot NeT bz @ FreeBSD dot org

6 8

[ipv6-wg@ripe.net] Draft Minutes (v1) IPv6 WG RIPE 50
by David Kessens 01 Jun '05

01 Jun '05

Please see below for our first cut of the minutes for our working group session during RIPE 50. Please let me know if you have any comments or corrections. The minutes will be declared final by Friday, June 10, 2005 if no significant changes (other than typo fixes and minor editorial issues) are found. David Kessens --- Draft Minutes (v1) IPv6 WG RIPE 50 When: Friday, 6 May 2005, 09:00 - 10:30 Where: C58, Clarion Hotel, Stockholm, SE Minute Taker: Adrian Bedford Quick Update from the RIPE NCC Regarding IPv6 Services (Ruud de Kooter - RIPE NCC) After the presentation, Ruud was asked if there were plans to integrate the IPv6 whois interface with the main whois. Currently the proxy cannot talk to the interface and this hampers some functionality (such as rate limiting). Ruud asked Shane Kerr to answer this. Shane explained that the RIPE NCC has a v6 proxy running over a v4 database backend. The RIPE NCC has been looking at how people use the database for two years and patterns suggest that now would be a good time to integrate native IPv6 into the server using /48 as a client identifier. Discussion on: Global IPv6 Routing Table Status (Gert Doering, Spacenet) Gert indicated that the Routing Working Group was discussing best practices for BGP filters. The overall thrust of the presentation was that v6 is growing smoothly with no big disasters so far. A question was asked about ratio comparison between the numbers of Autonomous Systems (AS) announced by IPv6 networks as opposed to v4. The current data suggests that there are 1.419 announced entries per originating AS for v6 and 10.308 for v4. A point was made that Deutsche Telecom (DT), who have one of the largest v6 allocations made to date would not announce their block as one AS, they would be more likely to split this and allocate blocks to different service providers for political reasons. Someone from DT commented that the company has the right to do this, though denied that the move was in any way politically motivated. Renumbering IPv6 Networks (Stig Venaas, Uninett & Tim Chown, University of Southampton) After the presentation, it was mentioned that we are assuming that customers of service providers will keep their space, yet ten years ago major providers did force renumbering onto customers. It was agreed that renumbering of IPv6 is, at least in principal, no different to what happened with v4. Developments/Initiatives Regarding IPv6 in the RIPE Region and Beyond - Francis Dupont: Point6 (http://www.point6.net/ - Gert Doering: A new maillist for discussion of ipv6 operational issues was started: http://lists.cluenet.de/mailman/listinfo/ipv6-ops/ After the announcement of the list, it was agreed that having a publicly available archive of this new list would be a good idea. Although it exists on the site that hosts the list, this is a personal website and for stability purposes, maybe a copy of the archive should be placed elsewhere. It was felt that the RIPE website was unsuitable for this. - Carlos Friacas: o IPv6 - Advanced Network Developments's first year in Portugal o 9K IPv6 Schools' Network Initiative o About the IPv6 Steering Committee Revisiting the /48 recommendation There was a lengthy discussion around RFC 3177. It was made clear that the discussion was not intended to reach final consensus on all the issues involved, but merely to serve as a way to get an idea about the general direction on where the community would like to go to make it possible to write up an initial proposal. Many agreed that the /48 is too much for many users for now and in the future. A discussion followed about how we should define a site with the suggestion that we need to accept that this can remain a grey area, which is not at all well defined right now. Discussion is needed around this topic, though the group tended to accept that it might not be easy to put down limits in black and white. Most people in the room felt that RFC3177 needs a revision. Most people felt that the /48 limit itself doesn't need to be changed, but most agreed that there is a need for a new category that falls somewhere between /48 and /64 for users that have a need for subnetting but for which a /48 seems excessive. This category would fit the mass consumermarket or (very) small office market, however, it was noted that we should base the category on technical needs of the user, not in economic terms like mass consumer market. It is exepected that an Internet Draft will be written that can be discussed at RIPE 51.

1 0

[ipv6-wg@ripe.net] Minutes from IPv6 WG - RIPE 49
by David Kessens 01 Jun '05

01 Jun '05

Please see below for the final minutes of our session during RIPE 49. The minutes are mostly the same as the v1 draft minutes that were posted earlier. Only differences are a typo fix and an improved summary of Jordi's presentation on 'Guidelines for ISPs on IPv6 Assignments to Customers'. I will sent out the draft minutes for our meeting during RIPE 50 in a couple of minutes. David Kessens --- Minutes from IPv6 WG - RIPE 49 RIPE 49 Manchester IPv6 Working Group Wednesday 22 September 2004 WG Chair David Kessens Scribe Adrian Bedford (RIPE NCC) Attendees: 121 Administration and Agenda Presented by David Kessens Adrian Bedford of the RIPE NCC was appointed scribe. Agenda was agreed Minutes from both RIPE 47 and 48 were approved. Update from the RIPE NCC Presented by Andrei Robachevsky, RIPE NCC As of August, k-root is IPv6 enabled The new version of the RIPE website (to be launched shortly) will also be IPv6 enabled. An Overview of the Global IPv6 Routing Table Presented by Gert Doering (Spacenet) A number of anomalies have been observed in the Global IPv6 Routing table, though in most cases these could be explained by traffic engineering, internal transitions and in one case a prefix leak. One person used the Jabber Remote Feedback system to ask why IXPs are announcing /48s with no export. It was explained that, for example, AMSIX doesn't want it's neighbors to make it's prefix globally visible. It remains a contentious issue as to whether something should be globally visible. What are the Deployment Plans Behind the Larger IPv6 Allocations? Presented by: Mikael Lind (TeliaSonera) David Kessens asked whether there were plans to allocate /64s for mobile use. Mikael confirmed that this was likely and that TeliaSonera was also looking at /48s for this use. Kurtis Lindqvist asked whether TeliaSonera has been talking with vendors to co-ordinate hardware and equipment. Mikael explained that this has not yet happened, as they feel right now there is little incentive to do so. David asked if Mikael would consider returning to a future RIPE Meeting to provide an update when TeliaSonera is further along in the development and deployment process, Mikael indicated that he would be happy to do so. Deployment Plans Behind Larger IPv6 Allocations Presented by Jordi Palet (Consulintel) Jordi will continue to keep the Working Group posted on feedback. He will also encourage non-respondents to contribute. Geant IPv6 Presented by Jordi Palet (Consulintel) Guidelines for ISPs on IPv6 Assignments to Customers Presented by Jordi Palet (Consulintel) This presentation provided an initial attempt on creating guidelines on how to interpret the official ipv6 assignment policy in practical terms. < BREAK > IPv6 Glue for the Root Nameservers Presented by Johan Ihrén (Autonomica) Iljitsch van Beijnum asked how long the process is likely to take. Johan commented that it would depend on who was prepared to take on the task, Johan is willing to do this, though added that nothing is root server specific, if people argue that we need v6 glue now, then they should step up and offer to help out. There is no specific target date for this to happen, since testing needs to be completed. Randy Bush asked that since we tend to assume that we should upgrade when something breaks, how is he to know when something has in fact broken. How, as an operator will he actually know there is something wrong? Johan agreed with Randy, there needs to be an agreed process. This could mean that things may not happen universally at the same time. Much will depend on the level of ambition of those involved - the tasks could be more or less complex. Gert Doering commented that as we see ccTLDs adding v6 glue, they might already know some of the issues that could arise. He suggested that someone speak to the ccTLD Operators and find someone who is willing to help with testing. It was also suggested that prudence might be a good idea. If a ccTLD Operator decides to deploy v6 transitions to their nameservers, in the end it will be their own decision. A question was asked about whether anything could be done to improve I-root connectivity to Europe. Johan commented that this would fall into his domain and promised to provide more feedback on this by RIPE 50. Bill Manning commented that the group appeared to be talking as if testing had not in fact started, whereas people are doing tests, ccTLD Operators at already looking at the impacts of these tests. A timeframe was agreed at the last ISOC meeting to finish testing before their next report at the end of the year, enabling a report to made to ICANN. One obstacle here is that the last report to the ICANN board took six months to have approved. Doug Barton of ICANN commented that the process has been streamlined and lessons learned, it is felt that any such future report would receive a far quicker response. David Kessens asked if all root servers planned to have IPv6 connectivity at the same time. Johan replied that he thought it unlikely that everyone would approach this in the same way, adding that he felt diversity was not a bad thing. It would, he said, be likely that we would see some TLDs using basic glue with others holding off during the early phases. The day when we will see the v6 glue globally is still somewhere in the distant future as there are no specific plans at this point to co-ordinate such a deployment. To date eight have basic glue in place. David asked if this community could help and Johan said he felt it certainly could. There was discussion about asking the RIPE NCC for help, though Johan this was not necessary at this stage. From his point of view he did not feel the issue was who did this, rather thought needed to be given to identifying which tests are relevant. Time would need to be taken over inventory to find out what code is already out there. David rounded up the discussion by saying that volunteers would be welcomed, though this was not be taken as a specific action point. An Update on Multihoming in IPv6 Presented by Geoff Huston (APNIC) A comment was made that the slide detailing issues appeared to not consider the update rates for bindings and third party referrals. Geoff said that design decisions should avoid relying on DNS. Randy Bush commented that there had been some mission drift from multi-homing to mobility. A comment was made that if an opportunistic solution does not meet all the needs in the architecture, you are forced to use a structured identifier. Such an identifier would have to be unique to avoid collisions. This could give rise to a very complex registration infrastructure, which might be painful to administer. Geoff replied that uniqueness never comes cheap and needs enormous effort at a global level. This lead to a suggestion that it may be preferable to choose opportunistic styles. Geoff felt that this was down to personal taste; there were certainly benefits in avoiding global costs if set up problems could be avoided. A suggestion was made that we should consider developing a whole new protocol that could handle routing table growth, Geoff replied that it would be ideal to think that we could do both, groups do exist with experience in both fields. The IETF have working groups that work together to engineer solutions and ensure routing tables do not become too complex. Geoff agreed that many people do feel routing could provide a better answer Kuris Lindqvist pointed to work that had been done during an interim meeting in Santa Monica in June 2004, this looked at the architectural work done by Geoff and the 30-40 drafts submitted. The drafts were split into four or five categories and based upon these; a simple poll was carried out. In some categories, the authors withdrew their proposals and in others, there was little support. The consensus was that the architecture proposed by Geoff was the way the group wished to move forward at the moment. A speaker announced that he had prepared a draft on the use of PI to reduce pressures on routing tables, intended as a compromise and is accepting comments on this. A speaker added that during his presentation Geoff dismissed mobile IP for security and state management reasons. The IETF has already addressed security and state management could throw up more problems. He felt that having two solutions for these issues problem was unwise and we need to work towards a single solution for what seem like related problems. He felt that using security, as a reason to reject the architecture draft was a red herring. Geoff pointed out that this is not yet in the draft, though will be in the next version. Geoff agreed that some of the assumptions we made previously in mobile IPv6 now seem less secure. Security locators assume a stable home base - if it fails, everything else fails. Multi6 looks at the ‘homeless vagrants’ of the world that do not have a home all of the time and the home is not reachable. It then assumes that there is not a home locator that can always be contacted. The Multi6 team looked at the other approaches and aimed to learn from them. He remains unsure if this kind of binding works in mobile environments, it is far too early to make this assessment. David Kessens asked whether the group found updates from Multi6 useful, Kurtis suggested that it would be nice to have Geoff make a similar presentation in future Address Policy Working Group Sessions too. Gert Doering suggested that as co-chair of the Address Policy Working Group, he would like to see the drafts on PI Addressing discussed through the Address Policy Mailing List. Moonv6 Update Presented by Jim Bound Kurtis Lindqvist asked whether security testing had been carried out in an isolated test environment. Jim clarified that filters had only been tested on IPv6 ports, addresses and protocols. He added that it showed there was confusion. Just because dual stack is not a security problem, IPv6 in itself is not an insecure protocol. Intrusion detection needs to be included into work. Kurtis asked what stage of the evolution process the project had reached. Jim replied that in this respect, research was still quite basic. Deploying 5,000 IPv6 Sites - XTEC Presented by Jordi Palet (Consulintel) Input to RIPE NCC Activity Plan Presented by David Kessens David asked if the group had any items to include in the 2005 RIPE NCC Activity Plan. There was no response.

1 0

[ipv6-wg@ripe.net] Re: [ncc-services-wg] RIPE 50 Report
by Daniel Roesen 25 May '05

25 May '05

On Tue, May 24, 2005 at 02:09:07PM +0200, Nick Hyrka wrote: > IPV6 WORKING GROUP > > - It was decided that the RIPE Whois Database will continue to include more > services that run in native IPv6. Hm? > - It was noted that the IPv6 Working Group needs to decide where it wishes > to house the new, operational IPv6 mailing list. What are you talking about here? What "new, operational IPv6 mailing list" are you referring to? What has the RIPE IPv6 Working Group to do with it? I have some suspicion that this is a reference to Ulrich Kiermayr's comment? Please clarify. Regards, Daniel -- CLUE-RIPE -- Jabber: dr(a)cluenet.de -- dr@IRCnet -- PGP: 0xA85C8AA0

4 5

[ipv6-wg@ripe.net] Fwd: [apops] APNIC New IPv6 address block(s)
by leo vegoda 24 May '05

24 May '05

Begin forwarded message: > From: John Tran <john(a)apnic.net> > Date: May 24, 2005 9:29:05 am GMT+02:00 > To: apnic-announce(a)lists.apnic.net > Cc: apops(a)apops.net > Subject: [apops] APNIC New IPv6 address block(s) > Message-Id: <Pine.LNX.4.61.0505231059340.10511(a)durian.apnic.net> > > > > Dear colleagues > > APNIC received the following IPv6 address block from IANA recently > and will be making allocations from this range in the near future. > > 2400:0000::/19 > > This announcement is being made for the information of the Internet > community, and so that network configurations such as routing filters > may be updated as appropriate. > > For more information on the resources administered by APNIC, see: > > http://www.apnic.net/db/ranges.html > > For information on the minimum allocation sizes within address ranges > administered by APNIC, see: > > http://www.apnic.net/db/min-alloc.html > > > Kind regards > > Son > > ____________________________________________________________________ > Resources Services Manager <Son(a)apnic.net> > Asia Pacific Network Information Centre phone: +61 7 3858 3100 > http://www.apnic.net fax: +61 7 3858 3199 > Helpdesk phone: +61 7 3858 3188 > email: > helpdesk(a)apnic.net > Please send Internet Resource Requests to <hostmaster(a)apnic.net> > _____________________________________________________________________ > > _______________________________________________ > apops mailing list > apops(a)apops.net > http://mailman.apnic.net/mailman/listinfo/apops > Website: www.apops.net > -- leo vegoda Registration Services Manager RIPE NCC

1 0

[ipv6-wg@ripe.net] Who provides v6 uplink via tunnel ?
by Kurt Jaeger 22 May '05

22 May '05

Hello, As none of our uplinks currently provides v6 support, I'm looking for some site that would be willing to route our /32 and tunnel it to some endpoint via v4. We're located in southern germany (Stuttgart). -- MfG/Best regards, Kurt Jaeger 15 years to go ! LF.net GmbH fon +49 711 90074-23 pi(a)LF.net Ruppmannstr. 27 fax +49 711 90074-33 D-70565 Stuttgart mob +49 171 3101372

5 8

[ipv6-wg@ripe.net] What is a site?
by ＠UUNET SE Ip 11 May '05

11 May '05

Hi. Following the discussion about /48 boundaries I'd like a better definition of what a site is. My definition of an end-user site is the office where we (MCI/UUNET) install a circuit. This could be a large office or a small bransch office or anything in between. Each office is handled separately and they request IPv4 addresses per office. Adopting this to IPv6 it would mean that each office would get a /48. This is too much for many of them. Approx. 80-90% of our sites request 32 IP-addresses or less and most likely only subnet it 2 or 4 times if they ever subnet it. Furthermore we never get a complete network design for all branch offices if the customer is transnational or even national so we can't really assign a /48 to the customer for him to subnet among his offices. We can of course change our procedures to accomomodate thisbut this will make things a lot more difficult for us as a LIR and also the customer. We can't decide how many addresses to assign to a customer based on size or revenue. What I want is a clear definition of what a site is by having more catagories. but I don't want a floating boundary as catagories do simplyfies things. I also include my suggestions based on where I'm coming from :-) /60 for home networks (16 networks) /56 for enterprises (small/medium) (256 networks) /48 for large enterprises (65000 networks) /47 or more for "very large subscribers" /64 for mobile phones (w/ bluetooth or 802.11b) /128 for dialup PC Please note that I'm not a routing expert nor am I a experienced in the complete IPv6 concept with mobil users and consumers goods etc. My main area is IP-address administration. I administrate IP-networks for MCI/UUNET in almost all of Europe except DE, AT and CH. I haven't read the RFC's related to this so my mind is wide open ;-) Best regards Patrick Arkley Supervisor IP/DNS-team SKSC/SKRC MCI - Powered by UUNET Armégatan 38 S-171 04 Solna, Sweden Web: www.se.mci.com <http://www.se.mci.com/> Phone direct: +46 (0)8 5661 7075 Fax: +46 (0)8 5661 7236 Mobile: +46 (0)733 11 20 75 VNET: 915-7075 E-mail: patrick.arkley(a)se.mci.com <mailto:patrick.arkley@se.mci.com> Initial Call team [VPN]: +46 (0)8 5661 7899 or emea-ip-vpn-initial-call(a)se.mci.com <mailto:emea-ip-vpn-initial-call@se.mci.com> Registry: +46 (0)8 5661 7454 or registry(a)se.uu.net <mailto:registry@se.uu.net> <mailto:registry@se.uu.net> IP-team: +46 (0)8 5661 7629 or ip(a)se.uu.net <mailto:ip@se.uu.net>

13 44

[ipv6-wg@ripe.net] Re: [ipv6-wg@localhost] Re: What is a site?
by Erics 10 May '05

10 May '05

Hi all, please, let´s get back to the main question - what is a site ? I think we need a definition for this, because we find a lot of different means (eg. endsite, single network, customer, subscriber) in the IPv6 allocation and assignment policy and the RFC3177. Lets view a concrete example : A customer of us, a shipping company has ten brances at ten different locations. And he is running a seperated network behind our acces router on each loctaion. >From our point of view and understanding the RFC3177 each location is a site and we assign a /XX to each branch. Do you agree on this, or are there different opinions ? Kind regards Eric Schmidt LIR de.telekom

2 1

RE: [ipv6-wg@ripe.net] Re: [ipv6-wg@localhost] Re: What is a site ?
by Patrick Arkley 10 May '05

10 May '05

That was indeed the original question. In my scenario, the same as yours, I would assign a /48 to each branch office. Too many subnets I would say. However, if the customer from a central point asked for a range for all his branch offices I wouldn't hesitate in assigning one /48 for him/her to use wherever appropriate. The problem from an adminitrative side is that we rarely get the customer to fill out a network request for all sites but rather request ranges per site as they are up for installation. Rgds Patrick Arkley Supervisor IP/DNS SKSC/SKRC Phone direct: +46 (0)8 5661 7075 Fax: +46 (0)8 5661 7236 Mobile: +46 (0)733 11 20 75 VNET: 915-7075 E-mail: patrick.arkley(a)se.mci.com Initial Call team [VPN]: +46 (0)8 5661 7899 or emea-ip-vpn-inital-call(a)se.mci.com Registry: +46 (0)8 5661 7454 or registry(a)se.uu.net IP-team: +46 (0)8 5661 7629 or ip(a)se.uu.net -----Original Message----- From: Erics [mailto:EricS@t-com.net] Sent: den 10 maj 2005 11:37 To: ipv6-wg(a)ripe.net Subject: [ipv6-wg(a)ripe.net] Re: [ipv6-wg@localhost] Re: What is a site? Hi all, please, let´s get back to the main question - what is a site ? I think we need a definition for this, because we find a lot of different means (eg. endsite, single network, customer, subscriber) in the IPv6 allocation and assignment policy and the RFC3177. Lets view a concrete example : A customer of us, a shipping company has ten brances at ten different locations. And he is running a seperated network behind our acces router on each loctaion. >From our point of view and understanding the RFC3177 each location is a site and we assign a /XX to each branch. Do you agree on this, or are there different opinions ? Kind regards Eric Schmidt LIR de.telekom

1 0

[ipv6-wg@ripe.net] HD ratio issues and revisiting /48
by Iljitsch van Beijnum 05 May '05

05 May '05

Consider this my input for tomorrow's discussion. I'm far away but I'll be with you guys in spirit. (And watching!) Ok, this message is becoming rather long. Skip to the end for the practical stuff if necessary. I believe Geoff's point with the presentation yesterday was "with only 45 bits to play with and considering the HD ratio IPv6 address space isn't as unlimited as it once seemed". The whole idea behind the HD ratio is that the more you have, the more you waste. And "more" isn't just more in absolute terms, but also in relative terms. Essentially an HD ratio of X percent says you'll waste X percent of your address bits. To some degree, this makes sense. If you have a flat numbering plan you need N numbers plus some wiggle room. But if you introduce a strictly hierarchical addressing plan, you need more wiggle room. For instance phone numbers we have in most European countries have area codes for cities and then subscriber numbers within each city. When phone numbers in Amsterdam run out, they can't borrow numbers from The Hague. So not only do we need wiggle room for both aggregation levels (national and city), but in phone numbers we also lose some serious number capacity because the boundary isn't very flexible. In the Netherlands even the smallest area codes hold 6 digit subscriber numbers. This eats up more numbers than when you can give the smallest area codes 5 digit numbers, or better yet, use 4.5 - 7 digit as needed. Now the interesting thing is that the HD ratio RFC (RFC 3194) is for a large part based on experiences with French phone numbers. It's easy to see that even though Alain Durand and Christian Huitema are obviously on to something, there must be a problem somewhere, as they write that the practical maximum (= HD ratio of 87%) utilization for IPv4 is 240 million addresses. (Actually that would be 211 million, they didn't compensate for class D and E and other unusable address space.) However, according to the most recent ISC host count were now at 317 million (= hosts present in the reverse DNS) while only 2.5 billion out of 3.7 billion usable IPv4 addresses have been allocated by IANA. So the current HD ratio for usable non-reserved IPv4 space is 90.5%, and this includes legacy space, so if we were to look at RIR space only, this would certainly be even higher. If we apply the same HD ratio to the remaining 1.2 billion addresses, we should be able to grow to 480 million IPv4 addresses in use without even taking legacy space issues into account. The HD ratio also doesn't account for the nature of the number hierarchy. In phone numbers or classful IPv4, you can't move the aggregation boundaries around. But with CIDR and VLSM (variable length subnet masks, at one point a great innovation), this is not a problem. So if you have 8 locations that need 100 - 800 addresses respectively, you don't need 8 x 1024 = 8192 addresses, but 128 + 256 + 3 x 512 + 3 x 1024 = 4992 addresses, or even 1, 2, 3, 4, 4, 5, 6, 7 * 128 = 4096 addresses. This will of course push some of the pain up the next aggregation level, as rather than a single route there will be 8 to 32 individual routes. When we cross over to inter-domain routing, this is clearly harmful since we are currently unable to aggregate in the interdomain routing space unless aggregation was already done (or at least possible) at the source AS. But for interior routing, this really isn't much of a problem. Even when routing a billion (2^30) /48s, that would boil down to: 2 levels, strict hierarchy: 2 * 2^15 routes 3 levels, strict hierarchy: 3 * 2^10 routes This is all very doable with today's hardware. Even if we assume a size difference of a factor 64 between the largest and the smallest entities in a loose hierarchy (ie. if Mexico City is your largest entity, the smallest would be a city with less than half a million people) that's 2^15 + 2^21 with two levels and 2^10 + 2^16 with three levels. Conclusion: you don't _have_ to lose one bit out of five as the allocated address space gets bigger. This also applies to interdomain routing. If you want to number 4 billion end-sites, Geoff's presentation lists a /8 as the prefix length of choice. But 4 billion also happens to be 256 * 16 million, and for 16 million we only need a /18. So 256 of those would be a / 10. We just saved 2 bits but now we have 256 /18s in the global routing table rather than one /8. But even someone who gets as much flack for not trusting the router vendors to come up with bigger boxes as me can appreciate that routing a larger number of /18s isn't going to kill BGP (there can only be 32 - 256 thousand /18 routes anyway). (With apologies for the thousand = 2^10 and million = 2^20.) !-!-!-! So where does this land us in the "create more wiggle room" debate? I think it's pretty clear that there is no immediate danger if we don't. On the other hand, if we can save some bits without real pain, why not do it? Going from /48 to /56 as a general recommendation would be one way to do that. Another way would be to keep the /48 for people who want it, but move the /64 to /60. It's practically impossible to utilize /128s in IPv6 (it can be done but it's just too painful and awkward). So: Now: /64 or /48 /56 idea: /64 or /56 /60 idea: /60 or /48 Giving everyone who needs address space a /60 rather than a /64 has the advantage that there is always room for a router. If my cell phone gets a /64, that's nice, but it means the phone has to bridge or proxy ND or some such. With a /60 I can have a subnet of my own for the bluetooth link, and another for the GPRS/UMTS link. Obviously this will increase the address consumption for people who would otherwise have gotten a /64, but even if there are many of those, that doesn't add up too badly. However, 95% of the people who would have gotten a /48 and 99% of the people who would have taken a / 56 would also be happy with a /60, and this would save a lot of address space. I also believe /60 vs /48 will make ISPs happier, as they're used to making SOHO/BigCu$tomer distinctions today and /60 vs /48 would pretty naturally fall into this. However, I stronly believe that EVERYONE who wants one and has an ISP who is reasonable should get a /48 no questions asked for the exact reasons that we have the "let them eat /48s" rule today. I would also suggest that all individual /48s would be registered in RIR databases.

1 0