July 2002 - lir-wg - mailman.ripe.net

Draft minutes from lir-wg at RIPE 41
by Hans Petter Holen 10 Sep '02

10 Sep '02

Dear WG, The RIPE NCC have just finished the draft minutes and published them on the RIPE web site under: http://www.ripe.net/ripe/wg/lir/r41-minutes.html RIPE Meeting: 41 Working Group: LIR Status: 1st Draft Revision Number: 1 Please mail comments/suggestions on: content to the Chair of the working group. format to webmaster(a)ripe.net. ---------------------------------------------------------------------------- --- Draft Minutes of the LIR Working Group session of the RIPE 41 Meeting 14-18 January 2002, at the Krasnapolsky Hotel, Amsterdam, Netherlands. ----------------------------------- Chair: James Aldridge Vice-Chair: Denesh Bhabuta Scribe: Jeannette Decades ----------------------------------- Agenda: A. Admin - scribe, participant list, charter, mailing-lists B. Agenda bashing C. RIPE 39+40 (minutes & actions) D. Report from the RIPE NCC Hostmasters E. RIPE 185 ("European Internet Registry Policies and Procedures") F. Change of policy for Portable address space G. RFC 2050 ("European Internet Registry Policies and Procedures") revision H. Report from The Address Council Y. Any Other Business Z. Open Microphone. Denesh Bhabuta (DB): Hans Petter Holen couldn't make it. So the meeting will have to do with James Aldridge and myself, Denesh Bhabuta. We want to set a ground rule: One of the complaints from the past was that the LIR WG has been fed with too many discussions. Discussions are supposed to happen on the mailing lists and policy decisions are to be made at the meeting. James Aldridge (JA): The agenda was sent out yesterday to the mailing list. ======================== A: Scribe. ======================== DB: Could we have a non-RIPE staff scribe please? No reactions. A RIPE staff member is scribe. ======================== B: Agenda bashing ======================== JA: Are there any additional items for the agenda? No reactions. ======================== C: RIPE 39 and RIPE 40 minutes and actions. ======================== JA: The minutes are on the website since November we think. Actions as it appears on the website, few of them are very old actions from RIPE 35, 36 and 38. Most of these had something happened since the last RIPE meeting. A lot of it is ongoing work. The 38.5 LIR-Allocated, my original suggestion, is now LIR-Partitioned and some email is sent out this week. It is now in the test database and will go live real soon now. Working group chair roles is working fine since Hans Petter Holen is not here and I am. Policy action undivided address space action on the working group to participate on the RFC 2050 is in progress. ======================= D: Report from the RIPE NCC Hostmasters ======================== Presentation of Sabrina Waschke can be found at: http://www.ripe.net/ripe/meetings/archive/ripe-41/presentations/lir-rsreport /index.html Some additional comments from Sabrina on the slides: RS staffing. We have now 5 teams. Three new hostmasters are starting in January. At this point we hope to be fully staffed and can cope with the workload. And as you can see, you have probably seen it before, we also work on future growth. (Sabrina shows her big belly). Number of requests in hostmaster(a)ripe.net. In 2000, we see a peak but now the graph shows a steady growth. We introduced last year the two-day-turnaround time for ongoing tickets. That meant that when your request was picked up out of the wait queue, a hostmaster dealt with it, you replied, you had to wait for 48 hours before you got a response from us again. That means that correctly filled out requests get a faster approval. The special lir-help(a)ripe.net mailbox is not holding up normal requests anymore. We still apply the 'approve but educate' approach. Even not perfect requests got approved and we gave tips and information for improvements to get a faster approval next time. Wait queue. The last 6 months the wait queue is low. The last 2 months it is a bit higher, due to Dutch weather (sickness) and Christmas holidays. Total LIRs per region. The growth is slowing down. We don't expect too many new LIRs. IPv4 Allocation distribution. LIRs that are working all over Europe have the country code EU. Policy development. A message was sent out to the LIR WG concerning the /20 criteria. At RIPE 40 the decision about the Assignment Window (AW) for infrastructure for LIRs was decided upon and it was introduced in December 2001. Already 250 objects have been created in the database with the INFRA-AW extension. Not all of them are valid. Some LIRs have tried to use this to clean up the database. RS services are working on a document that tries to cover everything concerning mergers, take-overs and closures of LIRs. IPv6. The assignments for Exchange Points have been introduced in August 2001. E: RIPE 185 ("European Internet Registry Policies and Procedures") ======================== Nurani Nimpuno (NN): We encourage everyone to look at the document and to give comments. We're speaking about the new policy draft document for IPv4 policies. The document has been out on the web for 5 months and there was very little activity. The couple of comments that have been made have been published on the LIR WG mailing list. JA: I'm not sure how much more we can do. There hasn't been very much discussion on the mailing list about this. My opinion is this, more profitable work is done on the mailing list. Then we can come to the RIPE meetings for final decisions, rather than having long drawn-out arguments and conversations here. But are there any brief comments from anyone here? Because we're actually ahead of schedule and I'm trying to think of anything to fit in a bit of time. I encourage everyone to join the mailing list and to participate in the revision process. The email address is ripe-185bis(a)ripe.net. ======================== F: Change of policy for Portable address space ======================== JA: The question is if people are happy with the new policy? No reaction. DB: Are people aware of the policy change? It's about the requirements of a /22 immediate or the previously use of a /22 to become a new LIR. This was voted on at the last RIPE meeting. Any thoughts? Are people finding it difficult with their customers who want to be mainly multi-homed? Comments? No? We will continue this discussion on the mailing list. ======================== G: RFC 2050 ("European Internet Registry Policies and Procedures") revision ======================== JA: The agenda on the web was wrong about this one. This should be the RFC 2050, the global Internet Registries guidelines, which is being revised and prolong sides the European policy. Again joining the mailing list for participation and discussion is appreciated. Are there any comments for now? Mark Mcfadden (mcf(a)uwm.edu): I'll give a brief update. There is a mailing list available through ARIN for the global RFC 2050 discussion. Part of that working group, there are also some milestones. We're a little bit behind on those because of some things that happened at the end of last year. But we can start see there again is some movement forward. So I encourage anyone who is interested in the discussion of the revision of RFC 2050 to make sure that you're part of that mailing list. Mailing list address: rfc2050-WG(a)arin.net JA: Any questions, comments, anything? Comment from ?: I think that there are more hostmasters here are than LIRs. ======================== H: Report from The Address Council ======================== Barbara Roseman gave a report from the Address Council. Slides can be found at: ======================== Y: Any Other Business ======================== UUNET: There has been a discussion on the mailing list about PI - multihomed issues. The question raised if multi-homed is always needed or not. We have customers that need multi-homing. But not all providers will route small address blocks. JA: I don't know if there's anything that the LIR WG can do to force providers to route any particular address space. I would be inclined, rather than a policy change our of LIRs, have the Routing WG work on this issue and come up with a 'best practice or something'. DB: We can talkabout 'best practice' but there are commercial realities behind this as well. A lot of ISPs have customer requirements to fulfill. You're not the only person who has spoken about this issue with us and it has mainly come up since the last policy change. This was about the minimum requirement of a /22 before we can get address space. Are there any suggestions? UUNET: As far as I can see, the company that I work for, which is UUNET, is thinking about(at least the hostmaster is, which I am not) having a policy where you accept other provider's PA space instead of trying to go the PI route. This is current thinking, this is no decision, but the problem is we don't have a consistent policy across our AS and we'd like to implement one. Before we implement it we'd like to have consistent policy within RIPE. The problem is that whenever we get the discussion, everybody starts jumping and shouting saying: "Multi-homing". Then the discussion dies because multi-homing is the issue. But before we solve that problem we would like to have something implemented that is workable right now instead of waiting until we solve the mult-homing discussion, which seems to be without a logical solution. DB: My suggestion is that we take this up with the Routing WG as well and maybe we can make this into an action item from this meeting. UUNET: I didn't raise it in the Routing WG yesterday, thinking that was more appropriate to the LIR WG. I guess there were some initiatives in the community, but they all died. So I wonder if there's anybody else that feels like this, something we need to resolve rather sooner than later. DB: Any other comments? No reaction. UUNET: I guess there's too much RIPE NCC people here as previously discussed. DB: We'll set it as an action point. JA: This is the first new action point from this RIPE meeting. Any other questions, comments? I think this is about it for this morning. Please participate on the mailing lists. It does make it so much easier if there's been discussion before RIPE meetings. We can then reach final decisions at these WG-meetings.

2 1

AS Number Policy - continued
by leo vegoda 02 Aug '02

02 Aug '02

Dear Colleagues, Thank you for your comments on my initial questions and the wide ranging discussion that took place. I would like to continue this discussion by asking two more questions: 1. Should the RIPE NCC check whether Autonomous System Numbers it has assigned become multi-homed within six months? If you would like us to do this, the RIPE NCC proposes to do so by looking at routing policies registered in the Routing Registry. 2. Should the policy be changed so that AS numbers not in use can be reclaimed? If there is consensus on this we propose to reclaim AS numbers of networks not multi-homed after six months. Kind regards, -- leo vegoda RIPE NCC Registration Services

11 23

PGP support
by Christakis Christophorou 30 Jul '02

30 Jul '02

Dear colleagues, I would like to implement a PGP authentication scheme with RIPE. I would appreciate any help including suggestions about any plugin to be used with an existing email client software (outlook express etc) or any completely independent software. I have tried PGP Freeware with success but keys state that this version is for "non commercial use" and Network Associates does not support it any more. I have also tried GnuPG but without any success (keys are created with 0 bytes). Thanks Chris

1 0

Re: New draft: "New Values of the 'status:' attribute for inet6num Objects"
by Wilfried Woeber, UniVie/ACOnet 26 Jul '02

26 Jul '02

Dear Leo and "Team", >The new draft is available from our web site at: > ><http://www.ripe.net/ripe/draft-documents/new-value-status.html> Looks good! >We would welcome comments on this new draft until Thursday, 1 August >2002. All comments and suggestions should be sent to <lir-wg(a)ripe.net>. Procedural question: How are we going to modify existing objects to conform to this proposal? Right now the template query returns: bash-2.05a$ wripe -t inet6num [ ... ] inet6num: [mandatory] [single] [primary/look-up key] [ ... ] status: [generated] [single] [ ] [ ... ] and it seems to be "optional" for the (new) ALLOCATED-BY-RIR case: bash-2.05a$ wripe -r 2001:0628::/35 [ ... ] inet6num: 2001:0628::/35 netname: AT-ACONET-19990920 descr: ACOnet Sub-TLA block country: AT admin-c: WW144 tech-c: WW144 tech-c: WK42 tech-c: CP8-RIPE notify: Domain-Admin(a)UniVie.ac.at mnt-by: RIPE-NCC-HM-MNT mnt-lower: ACONET-LIR-MNT changed: hostmaster(a)ripe.net 19990920 changed: hostmaster(a)ripe.net 20011019 source: RIPE Is this going to change? E.g. to: status: [mandatory] [single] [ ] The wording used in the draft seems to indicate [optional] [single] [ ]. I think some clarification would be helpful. Thanks, cheers, Wilfried. _________________________________:_____________________________________ Wilfried Woeber : e-mail: Woeber(a)CC.UniVie.ac.at UniVie Computer Center - ACOnet : Tel: +43 1 4277 - 140 33 Universitaetsstrasse 7 : Fax: +43 1 4277 - 9 140 A-1010 Vienna, Austria, Europe : RIPE-DB: WW144, PGP keyID 0xF0ACB369 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

3 2

New draft: "New Values of the 'status:' attribute for inet6num Objects"
by leo vegoda 25 Jul '02

25 Jul '02

Dear Colleagues, Thank you for the comments we have received on the draft document "New Values of the 'status:' attribute for inet6num Objects". We have revised it to incorporate the suggestions you have made. The meaning of the labels has been made clearer, in line with the comments received. ALLOCATED-BY-RIR - For allocations made by an RIR to an LIR. ALLOCATED-BY-LIR - For allocations made by an LIR or an LIR's downstream customer to another downstream organisation. ASSIGNED - For assignments made to End User sites The new draft is available from our web site at: <http://www.ripe.net/ripe/draft-documents/new-value-status.html> We would welcome comments on this new draft until Thursday, 1 August 2002. All comments and suggestions should be sent to <lir-wg(a)ripe.net>. Regards, -- leo vegoda RIPE NCC, Registration Services Assistant Manager

1 0

Mntners with only MAIL-FROM auth
by Andrei Robachevsky 25 Jul '02

25 Jul '02

Attention owners of mntner objects using _only_ MAIL-FROM authentication method. Dear Colleagues, We recently sent out a message to all contacts for maintainers that still used MAIL-FROM as the ONLY means of authentication. This method is no longer valid and no updates can be authorised if any of the maintainers checked are in this situation. The message pointed to the proposed procedure on how to fix such situation (http://www.ripe.net/db/mailfrom.html). The RIPE-DBM support team has been overwhelmed with fax requests to update these maintainers. At the moment we are not able to respond to these requests as quickly as we would like to do. We have found it necessary to give a higher priority to other support requests. Therefore the response time for updates to maintainer authentication may be 5 to 10 working days. Please do not send repeat faxes unless we ask for them and do not send e-mails asking for a status update on the changes. This only slows the whole process down. We will reply to each request as soon as we make the change or find that we need additional information. This message will also be sent to all contacts for maintainers that still use MAIL-FROM as the ONLY means of authentication. Please bear with us on this and we apologise for the delay. Regards, Andrei Robachevsky DB Group Manager RIPE NCC

1 0

Question to a multihomed PA-space connection
by "Thoß, Janko" 25 Jul '02

25 Jul '02

Hello, I have "little" question (but notice that I'm not a guru in BGP-routing and RIPE-problems). We want to connect a customer which don't want to register as a new AS. But he wants a connect to 2 provivers (different AS's). We have tried to get PI-space from RIPE but the automaticly response told us the following: -------- Quote -------- 1. Why is PI space required rather than PA space? - If PI is requested for multi-homing please explain why the second provider cannot route PA space as a more specific route(with the PA block holder adding a more specific route too). - The same policies apply to PI space as PA. Administrative ease and/or routing are not reasons for PI. Please read the information about PA/PI from the RIPE Document, IPv4 Address Allocation and Assignment Policies in the RIPE NCC Service Region, Sections 5.1.5 and 5.1.6: -------- Unquote -------- Can anyone tell me what is to do for that the second provider is able to announce the /24-PA from the first provider (to emulate a multihomed connection) or is it only possible to use the second provider for backup with the same space? wait hopefully for tips or hints regards, Janko Thoß -- envia.tel GmbH Projektmanager Leipziger Str. 116 b 04425 Taucha Tel.: +49 (0) 341 / 120 7069 Fax: +49 (0) 341 / 120 7052 Mobil: +49 (0) 173 / 379 0048 eMail: janko.thoss(a)envia-tel.de Web: http://www.envia-tel.de

2 1

join two objects
by Mohamed Ibrahem 24 Jul '02

24 Jul '02

Dear all I have some objects for the same customer and I want to merge them in one object. Please advise THANX Mohamed Ibrahim Ali Data Communication ENG

1 0

RE: Help Again Please !
by Razavi (Support) 23 Jul '02

23 Jul '02

I know about this site and i have some documents, but the problem is they dont help and dont give enough information, even i have training course booklet but its not clear ! Regards, --Cathy Razavi. -----Original Message----- From: Stephane Bortzmeyer [mailto:bortzmeyer@gitoyen.net] Sent: Thursday, July 18, 2002 5:00 PM To: Razavi (Support) Cc: bortzmeyer(a)gitoyen.net Subject: Re: Help Again Please ! On Mon, Jul 15, 2002 at 03:20:50PM +0430, Razavi (Support) <krazavi(a)dpimail.net> wrote a message of 9 lines which said: > How can i change our Database or how can i have access to our DB ? i want > to change the information about technical person and get a new Nichhandle ! Read RIPE's documentation. www.ripe.net, it is well organized.

2 1

Re: NCC#2002069654 HMQUESTION
by Knowtion Hostmaster Role Account 23 Jul '02

23 Jul '02

> You are correct to say that you can not remove a maintainer from a <aut-num> > object without adding a new one. > > The attribute <mnt-by> in the <aut-num> object is mandatory, the AS number > AS9179 is part of the <as-block> from the RIPE NCC, and if correct the > maintainer <AS9179-MNT> is being maintained by yourselves. How can I phrase this better ? AS9179-MNT (me / my company) is the current but out-of-date maintainer on an object. Without deleting the object (as it is not relly mine) how can I cease being associated with it ? I (AS9179-MNT) no longer am the maintainer of various objects that really belong to the company X in the UK. My company has not been associated with them or provided services to them for sometime. Is it good and recommended practice therefore for me to just delete the objects - this may affect someones service ? cc'ed to lir-wg in case anyone has a better idea... Petr

5 5