- ipv6-wg - mailman.ripe.net

[ipv6-wg@ripe.net] Draft agenda (v1) for IPv6 wg RIPE 44
by David Kessens 11 Dec '02

11 Dec '02

Hi, Below follows the draft agenda (v1) for the ipv6 wg for RIPE44. This is the first version of the agenda so I very much welcome your input on other/more agenda topics. Please let me know if you have anything that you would like to have covered. Also, I am still looking for speakers on various topics, please let me know if you are willing/able to help out. As for: Action point on DK: to bring the issue of wg scheduling to the WG chair lunch, and try to find a solution. The first part (my part :-)) of this action point has been completed. The scheduling problems with the DNS and EIX wg have been brought to the attention of the schedulers and I hope that we will get a better schedule this time. As an experiment, in order to help to achieve this goal of less scheduling conflicts, I offered to shorten the ipv6 wg session to one 2 hour slot to help create more time in the schedule. Thanks, David K. --- Draft agenda (v1) for the IPv6 Working Group Meeting RIPE44 A. Administrative stuff - appointment of scribe - agenda bashing (David Kessens) B. Status of the 6bone (David Kessens) C. IPv6 version of TTM (Henk Uijterwaal) D. Input for the DNS wg: Action point 42.1: Investigate the CNAME and other solutions for v6-reverse delegation (RIPE NCC) IPv6 capable RR DNS servers - what is needed, what are the issues ?!? (RIPE NCC) Action point 42.2: Give an overview of 6-to-4 reverse delegation issues at RIPE 43 (RIPE NCC) F. tentative: IPv6 state of art in the italian educational and research community. (Gabriella Paolini) G. Global IPv6 routing table status (Gert Doering) H. Workable approaches for ipv6 multihoming in the absence of a better solution than the one we use in the ipv4 Internet (I am looking for speakers) I. Report(s) about *actual* v6 traffic volume as compared to v4? what's *real* out there, not what's on powerpoint? (I am looking for speakers) J. Developments/initiatives regarding IPv6 in the RIPE region and beyond (input from the audience) K. input to the RIPE NCC Activity Plan (input from the audience) Z. AOB ---

1 0

[ipv6-wg@ripe.net] Draft minutes (v1) of ipv6 wg session RIPE43
by David Kessens 11 Dec '02

11 Dec '02

Hi, Please see below for the first draft of the minutes of last meeting. Correction/comments are very much welcome. The minutes will be declared final if no major defects (other than typos etc.) are found by Friday, December 13th, 2002. David K. --- Draft minutes (v1) ipv6 wg RIPE 43 Rhodes, Greece 11 September 2002 A. Administrative stuff ======================= Agenda is available at: http://www.kessens.com/~david/presentations/ The scribe was appointed, and the agenda was agreed upon. B. Status of the 6bone (David Kessens (DK)) ====================== Presentation is available at: http://www.kessens.com/~david/presentations/ Per 20020911, there are now 1187 ipv6 sites in 58 different countries on the ipv6 Internet. C. IPv6 version of TTM (Henk Uijterwaal (HU)) ====================== Presentation is available at: http://www.ripe.net/ripe/meetings/archive/ripe-43/ presentations/ripe43-ipv6-ttm/ Wilfried Woeber (WW): With 6net hat - is possible to make preliminary measurements available before the end of the year? DK: How is this service supported from the commercial point of view? Right now, you have to pay for the test-box. HU: You still have to buy a test box, and we operate it for you. Service fee is currently 3000 EUR. Q: Do you get v4 measurements with that, too? A: In principle, yes. WW: Technical question: the ntp protocol is used to encode v4 from clock-host partners to prevent forming loops. Anyone here knows more? HU: forward me the emails from the discussion. DK: I like this new service. This can be a reason for me to order such a box, because v6 production is important for us, and I want to be able to measure the performance. D. Action point 42.1: ===================== Investigate the CNAME and other solutions for v6-reverse delegation E. IPv6 capable RR DNS servers ============================== What is needed, what are the issues ?!? (RIPE NCC) F. Action point 42.2: ===================== Give an overview of 6-to-4 reverse delegation issues at RIPE 43 (RIPE NCC) Andrei Robachevski, RIPE NCC covered topic D, E & F. In regards to D & E, a prototype was made; setups how to support them was tested; no final decision yet. Status: in progress. It will be accomplished before the next ripe meeting Mohsen Souissi (MS): Are you aware of the technical notes, published last week, about DNAME, by the IETF committee? This is only meant for the smooth transition from .int -> .arpa Things are moving very very slowly. There is no dns server for ip6.arpa. "ns.ripe.net" is not ipv6-capable. I have already suggested several times to IANA that AFNIC server is available to be a secondary for ip6.arpa, but I received no answer... DK: There is an agenda point in the AOB about the Activity Plan of the RIPE NCC. We need to add this in there. My suggestion is that the RIPE NCC members make it clear to the RIPE NCC and the RIPE NCC Board that we want RIPE NCC to give some real attention to this. G. Global IPv6 routing table status (Gert Doering (GD)) =================================== Presentation is available at: http://www.ripe.net/ripe/meetings/archive/ripe-43/ presentations/ripe43-ipv6-table/ Q: Do you see people changing over, or announcing both? A: It was possible to see everything, all combinations. Q: Are the filtering problems still there? A: Not really, but old prefix was still in the table, although I was not announcing it any more... Q: Do you see any difference between number of announcements vs number of allocations per region? A: This was not researched, it's a good idea for the next time. Bernard Tuy (BT): While discussing the filters - we are at the point when we do need some recommendations from the RIPE community. DK: 6bone/IETF docs exist. Do we need new ones? GD: They need re-doing, can not be accepted by default. DK: Do we need recommendations? GD: We need to at least keep an eye on this, and think about it. We also need to think about multihoming issue. Let's put it on the agenda for the next meeting. DK: Any opinions right now? GD: Just one customer is doing it - "announcing more specific prefix" ?? I find it a good solution. Francis Dupont (FD): That does not scale. It is too easy. ISPs do not want to buy GB of memory to support this. But there's no good alternative solution. GD: Multihoming needs to be done by BGP : either /32, or /48 that can get filtered. There are two groups busy with it in the IETF, official&dead, and another one... Proposal for the presentation on the next meeting - "Approaches to multihoming and why everything sucks". FD: "ipv6mh" is good place to discuss this. It's a closed mailing list. DK: Personally, I think the way Gert is doing it is a good way. DK: Anyone who is a good candidate to lead the discussion, present existing solutions and explain why they don't work, please volunteer for the next meeting. H. Report(s) ============ About *actual* v6 traffic volume as compared to v4? What's *real* out there, not what's on power-point? H.1 baseline measurement for ipv6.int queries ---------------------------------------------- (Terry Manderson, APNIC) [URL - Presentation missing!] Q: How up to date are your graphs, and can you see RIPE43 meeting influence? A: From last Friday, so can't see this meeting. Q: different considerations - DNS traffic as protocol, and v6 traffic after DNS queries. how to get v6 related RRs. unless you have v6 from the root, except the glue, you don't know where else to go. Q: Can you see where are the queries coming from? A: These were the baseline measurements, not analysis. But, most cam from Japan. MS:.jp & .fr have been supporting v6 in DNS for years. We are asking for glue records in root, but according to the policy that is not possible. We are again asking RIPE NCC to support our efforts to tell to the root - v6 is harmless. DK:In the Activity Plan input to the RIPE NCC, we should be making it clear that this is important to us, as members of the RIPE community. [coffee break] H.2 some snap-shot info ---------------------- (Wilfried Woeber) [URL - Presentation missing!] DK: For the next RIPE meeting we will be preparing dns/v6/sec workshop. H.3 Discussion -------------- Do we want some more activity under this topic - to cover it in depth; to have more data? e.g TTM with more data. Since this is a _working_ group, more work from the participants is appreciated... I. Developments/initiatives =========================== regarding IPv6 in the RIPE region and beyond (input from the audience) I.1 Bernard Tuy, Renater ------------------------ We are deploying M6bone. 40 players, 15 non-fr countries. If you are interested, you are more then welcome to join us. http://sem2.renater.fr/m6bone/ What do we want to achieve, what kind of SW is used... I.2 Robert Blechinger, CyberNet ------------------------------- Number of queries per day, on their DNS server: 28 6bone, 2 6to4, 336 RIPE NCC allocated: 211 own, 115 from one peer. I.3 Mohsen Souissi Ready to publish [...] to draft standard status. I.4. DK The same week that there is a RIPE meeting in Barcelona in May there is a IPv6 Forum in Madrid. We are trying to schedule that at least IPv6-wg does not overlap. MS: This might be an opportunity to test video-conferencing over IPv6. Z. AOB ====== Z.1. input to the RIPE NCC Activity Plan ----------------------------------------- Please let your representatives tell to the RIPE NCC management and Executive Board your requirements, so that they can be included in the Activiy Plan. Z.2. 6bone transition to the RIRs ---------------------------------- There has been a talk in the LIR-WG regarding that. 6bone addresses are currently allocated by Bob Fink, which is inconvenient as being a single point of failure, and is outside of the RIR structure. If interested, make sure to take part in the discussion on the lir-wg list & meetings. The progres of the co-ordination efforts will be reported on the next meeting. The most important are the technical issues - ip6.int & ip6.arpa delegations. They will be put on the agenda for the next meeting. Z.3. Other ---------- Info: Bernard Tuy: 29 & 30 October, IPv6 conference in Paris 28 October, IPv6 Tutorial by Mohsen Souissi, like the one done on Monday in Rhodes. Andrew Miehs, CyberNEt: ipv6-wg meeting should not be at the same time EIX!! One of the issues - we'd like to establish peering points, but I can't express this in the IEX-WG. DK: Other people asked for this too, many many times. The same - DNS-WG also has many v6 topics now, and is also scheduled at the same time! WW: Instead of complaining, it's better to put a proposal - how can we improve the scheduling? Action point on DK: to bring the issue to the WG chair lunch, and try to find a solution.

1 0

[ipv6-wg@ripe.net] RIPE NCC's DNS server available in IPv6
by Daniel Diaz 09 Dec '02

09 Dec '02

Dear all, We are glad to announce that since November 29, RIPE NCC's DNS service (ns.ripe.net) is also available in IPv6. This was an action on RIPE NCC for RIPE 43 and it is now accomplished. ns.ripe.net internet address = 193.0.0.193 ns.ripe.net IPv6 address = 2001:610:240:0:193::193 Best regards. -- Daniel.Diaz Operations Manager RIPE NCC

3 2

[ipv6-wg@ripe.net] Re: RIPE NCC's DNS server available in IPv6
by Daniel.Diaz 09 Dec '02

09 Dec '02

Dear Carlos, et al, (Apologies for duplicates). Carlos Friacas <cfriacas(a)fccn.pt> writes: * Hello, * * I have a question about this, as pointed out by one of my colleagues here: * * As the SOA record for ripe.net., the glue record for ns.ripe.net. on * <whatever>.gtld-servers.net is 193.0.0.193, right? * Shouldnt they have also an AAAA record for ns.ripe.net on the net. zone? * Is this scheduled/possible/requestable/desirable??? * Indeed having this IN AAAA record included in the parent zone for ripe.net. would be the most desirable. So far I ignore though if .net nameservers are V6 ready or not. We definitively will look forward getting this record included in our parent zone etc... still the root-zone is not V6 ready yet, so the tree can not be completed anyway. With best regards. -- Daniel.Diaz Operations Manager RIPE NCC

7 6

[ipv6-wg@ripe.net] Re: posting privileges
by David Kessens 09 Dec '02

09 Dec '02

Hi, As you might have noticed, our deadline for comments passed on Friday December 6. I received private comments from the following people: Robert Martin-Leghne Gert Doering Michel Py Fearghas McKay Marko Veelma They all agreed that posting privileges for Jim Fleming should be removed. We did not receive any comments on the maillist that we should not take this measure. I will immediately request the RIPE NCC to implement our request. Thanks for your support. David K. --- On Wed, Nov 27, 2002 at 02:49:32PM -0800, David Kessens wrote: > > Hi, > > As you all know, something called Jim Fleming has been sending > off-topic mails to our mailing list for a while. > > Earlier last month, I was intent on asking the RIPE NCC to revoke his > posting privileges. However, after some private consultations, I > decided not to ask for that just yet. > > Due to his continuous abuse on our mail list earlier this month, and > no signs of diminishing activity on other RIPE maillists, I would like > to propose that I ask the RIPE NCC to remove posting privileges for > Jim Fleming. > > I plan to request this on Friday December 6 if I don't receive input > from the community that this is not desirable. > > David K. > ---

1 0

Re: [ipv6-wg@ripe.net] Re: RIPE NCC's DNS server available in IPv6
by Daniel.Diaz 03 Dec '02

03 Dec '02

Carlos Friacas <cfriacas(a)fccn.pt> writes: * On Tue, 3 Dec 2002, Daniel Diaz wrote: * * * Hello again. * * > Francis Dupont <Francis.Dupont(a)enst-bretagne.fr> writes: * > * In your previous mail you wrote: * > * * > * Indeed having this IN AAAA record included in the parent zone for * > * ripe.net. would be the most desirable. So far I ignore though if .ne * t * > * nameservers are V6 ready or not. We definitively will look forward * > * getting this record included in our parent zone etc... still the * > * root-zone is not V6 ready yet, so the tree can not be completed * > * anyway. * > * * > * => to support AAAA RRs in the glue and to support IPv6 transport are * > * two different things. And what it is asked for is the first/easiest one * . * * Of course. My original question objective was trying to know if RIPE * wants to take this one more little step. Saying its easy, and not doing it * doesnt make things go forward... :-) We'll definitively will look forward getting this AAAA record for ns.ripe.net included in the .net zone. Regards. -- Daniel.Diaz RIPE NCC

1 0

[ipv6-wg@ripe.net] Re: RIPE NCC's DNS server available in IPv6
by Daniel.Diaz 03 Dec '02

03 Dec '02

apologies for duplicates. "Daniel.Diaz" <ddiaz(a)ripe.net> writes: * Indeed having this IN AAAA record included in the parent zone for ripe.net ^^^^^^^^^^ |||||||||| Sorry... this was meant to be 'ns.ripe.net' * would be the most desirable. So far I ignore though if .net nameservers are V6 * ready or not. * We definitively will look forward getting this record included in our * parent zone etc... still the root-zone is not V6 ready yet, so the tree can not * be completed anyway. -- Daniel.Diaz

1 0

[ipv6-wg@ripe.net] posting privileges
by David Kessens 27 Nov '02

27 Nov '02

Hi, As you all know, something called Jim Fleming has been sending off-topic mails to our mailing list for a while. Earlier last month, I was intent on asking the RIPE NCC to revoke his posting privileges. However, after some private consultations, I decided not to ask for that just yet. Due to his continuous abuse on our mail list earlier this month, and no signs of diminishing activity on other RIPE maillists, I would like to propose that I ask the RIPE NCC to remove posting privileges for Jim Fleming. I plan to request this on Friday December 6 if I don't receive input from the community that this is not desirable. David K. ---

1 0

[ipv6-wg@ripe.net] resolv problem
by Max Gargani 15 Nov '02

15 Nov '02

Hi all, maybe this is not the right place or maybe I wrong something but since RIPE NCC gave to my company the reverse delegation for the allocated /32 ip6.arpa, all protocols such ssh, ircd etc etc, can't resolve the addresses. >From named log I see ssh & co ask to resolve i.p.v.6.a.d.d.r.e.s.s.ip6.int but all RIRs allocations are ip6.arpa. Anyone with same problem? Thanks, Max

1 1

RE: [ipv6-wg@ripe.net] IPv6 TLAs for mobile operators
by Andreas.Schmid1＠swisscom.com 15 Nov '02

15 Nov '02

> > On Thu, 31 Oct 2002, Arash Ashouriha wrote: > > A single terminal or mobile device should get /127 and not /64 ! > > /64 is for 1 network. If the customer need more than 1 > Network, he may get /48. > > Where did you get that idea? > > Please have a look at RFC3314 (Recommendations for IPv6 in Third > Generation Partnership Project (3GPP) Standards) and RFC3177 > (IAB/IESG > Recommendations on IPv6 Address Allocations to Sites). RFC3314 says on page 14: "Because multiple IPv6 hosts may attach through a 3GPP handset, the IPv6 WG recommends that one or more /64 prefixes should be assigned to each primary PDP context." But I really don't want to concentrate on that discussion. I want to know if RIPE NCC accepts IPv6 TLA requests from mobile operators only having GPRS/UMTS/WLAN customers - and therefore not providing connectivity to organisations with a /48. -> if anybody can clarify this, I would be very happy! > > Of course, anyone can build the networks as they want, but > assigning /64's > to mobile customers is something that must be supported by RIR's. > > > > I am new to this list - hope I do not touch a topic that > has been discussed different times in here.... > > > > > > I wonder how mobile operators receive IPv6 address > allocations from RIPE NCC? The IPv6 policy paper > (http://www.ripe.net/ripe/docs/ipv6policy.html) includes 4 conditions: > > > a) be an LIR; > > > b) not be an end site; > > > c) plan to provide IPv6 connectivity to > organisations to which it will assign /48s, by advertising > that connectivity through its single aggregated address > allocation; and > > > d) have a plan for making at least 200 /48 > assignments to other organisations within two years. > > > > > > Mobile operators fulfill a) and b). But since mobile > operators will probably give /64 addresses to their users (a > single mobile terminal), they don't fulfill c) and d) - if > they are not a traditional ISP in parallel. > > > > > -> how does RIPE treat IPv6 TLA requests from pure mobile > operators? > > > > > > > > > > > thanks for any answer > > > Andreas > > > > > > PS: one might argue that the mobile operators will give a > /48 for each mobile terminal. But this would mean that a /32 > gives them only 65k mobile terminals to address! A large > mobile operator haveing more than 10 Mio customers would need > 2^24 addresses resulting in a /24! Is this what is intended > to be given to a mobile operator from RIPE? I don't think so.... > > > > > > > > -- > Pekka Savola "Tell me of difficulties surmounted, > Netcore Oy not those you stumble over and fall" > Systems. Networks. Security. -- Robert Jordan: A Crown of Swords > >

7 6