- ipv6-wg - mailman.ripe.net

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