These are unedited transcripts and may contain errors.
Next session at 4 p.m..
NCC Services Working Group.
18 April 2012, 4:00?5:30 p.m..
KURT LINDQVIST: So, this is the NCC Services Working Group and we are really tight on time, so can you please all sit down and be quiet or leave.
We are going to get started. So, my name is Kurtis Lindqvist, I am one of the co?chairs of the NCC Services Working Group and as we all know, this is your favourite Working Group and we just want to share the love with you.
We have a very tight agenda today, and if you haven't been here before, the NCC Services Working Group contains presentations that are part of the AGM to follow. If you haven't registered for the AGM and gotten your badge, now is the very last chance to do so, and please do so. And the way we are going to do this is that we might run a little bit over time, we don't know yet. If we do, there will be 20 minutes between the time that you have emptied this room and the AGM starts. That means that once eventually we are finished here, we really need your cooperation to empty the room as quickly as possible and it's going to be me and the the others, we are going to chase you out of the room physically. By the time we finish here. But please leave the room as soon as we are done here.
So, onto the agenda and the meeting. First of all, welcome. But don't stay too long. We have a scribe which is law a from the NCC and Athina from the NCC is doing the Jabber, thank you. When you come to the microphone, please state your name so we know who you are, and we have a proposed agenda which looks like as follows:
Do I realise that every year we do this I forgot to put approval of the last minutes in there, and I don't think there was any comments in the minutes posted, so I assume we can all approve the last minutes. Minutes approved.
And agenda for today then is the NCC update an Exel, followed by member services development from Andrew, v4 final distribution process, Andrea and some service for legacy holders by Niall and then Jochem de Ruig with cross boarder law enforcement and last Daniel on the TTM and DNSMON services and I am going to hold the speakers to their time because if we make this we'll be on time. But this is without questions. So with that we will start up and first up then is Axel.
AXEL PAWLIK: Good afternoon. Welcome. I don't know what they call my presentation the RIPE NCC update when there is so much more after this from the RIPE NCC and about the RIPE NCC. The very high level update, though, from me.
Just five or six slides.
We are going, we are still growing after 20 years, that's a good thing. We are all looking forward to seeing how this is going to develop over the next couple of one, two, three years. We do know that APNIC has reached their final /8 and they are still growing, so...
The main focus of our work, the top most priority at this particular time is the registry to make sure that we have a strong, correct, current, up to date, complete registry is consuming quite a bit of our manpower. And you know some of the activities that we are performing in this corner are not particularly new. Some of them are brand new. So, you won't be surprised to hear more about this a bit later, I think mainly in Andrew's talk. 2007?01, that tells us something, it's five years old. We are still implementing it, still working away at at it with some quite some success but it is a pile of work.
Contacting legacy space holders. We talked about this last time and will continue to do so here. We are doing this, we again, found quite a bunch of them and are talking, that's good.
Auditing, we used to do a couple of audits a year. These days it's a couple of hundred a year. Auditing you, our members.
Report form. To report things in the database that are not quite what they seem. To us, that's one of the new things that we have been doing that sparked quite a bit of activity also on various mailing lists. A good thing there.
Now, IANA: The poor chaps have run out of IPv4 addresses, that's a sad state of of affairs, and we want to fix that. We have said earlier that where we get legacy address space back and we do get some back, I have gotten some back already, we will return that to IANA for further allocation to the RIRs, or re allocation back to the five of us.
Speaking of RIRs, that of course needs a global policy for post exhaustion IPv4 allocation mechanisms by IANNA. That things has passed to all the RIR regions. Is has passed to the address Council. It has passed the first 21 days public comment period at ICANN, it's currently in the second, you can comment on the comments I think, and I expect that the ICANN board will ratify this thing quickly by e?mail probably over the next couple of days. So that looks good.
In terms of services. God, we are doing so much and we are offing so much and we have big, big website and it's great and shiny. But also I get lost on it occasionally. And we have not discussed this internally and thought we must make this a little bit more clear. And basically, and you'll hear much more about that later on, is that we focus on four areas, and the areas ?? well up on the screen ?? is let's see, the LIR portal: Those are ways you can maintain your resources in our structure. You can look at them, you can update stuff. So that's just basically yours.
Then we have RIPE stuff, that's information about all resources and anybody can access that and look at them so that's all of the resources.
We have RIPE Labs, that's not entirely new. You have seen that before. Those are stories we have out there, some of them about what we are doing, but also lots of things about the general surroundings. Not just focused on number resources, addresses and the like, but more. And you have seen the beginning, baby steps ?? well, big baby steps by now, of Atlas measuring stuff on the Internet. So these four areas we want to use as headlines basically for everything that we are doing that should fall into those. It makes it a little bit clearer.
A stakeholder survey. That was last year. We have done that and thank you very much to all of you again who have helped us with that who have given comments, who have filled that stuff in we have taken a lot of information out of that, that was the idea obviously, and we have numbers 54 key findings that affect the organisation.
We have looked at that and divided the tasks among ourselves according to what we are responsible for internally, along those lines of basically activities. And looking at that, we are making great progress there. Some of that stuff has been completed. 16 of those 54 items have been completed. 33 are currently in programme. A couple of them are considered complete, but are ongoing, of an ongoing nature and a couple of them we are still trying to understand and investigate and see what we can do with them.
And there is more about that of course in the other presentations to some more detail. But basically the message here is that we are doing something with the information you are giving us, and of course you are not getting away. We are already planning the next survey to be done next year.
Right. Other things that you know we have been doing. Certification was a really big topic at the last RIPE Meeting and general meeting. And you probably have seen Alex's presentation yesterday in the Plenary. We think we are pretty much on track in addressing the concerns that we have heard from our members from the greater community. If you have any further, please do talk to us. We want to make sure that you feel comfortable about this activity and service.
The RIPE NCC charging scheme taskforce. That was an interesting topic last time as well at the general meeting. We have set up a charging scheme taskforce, those great people were our membership have met a couple of times and delivered a report and we'll talk about that later on I think in the general meeting. Obviously the full presentation of what the outcome of that is, we'll see. There is police order and everything that came from there. Also we have a full presentation on that. So probably you'll see what we are doing. We strive for full transparency on those issues. Any developments we put on the website, write a short announcement about so that you are in the clear about what your organisation is doing here.
Outreach external relations. You see Paul here, you also probably know that I have sent him off to Dubai, not because I don't like him any more, it's just he likes the heat and, you know, the locals there, and he is very, very busy representing us and you there in the regions, not only in the Middle Eastern region, but also in the Russian and Ukranian areas. This is something that we still see areas that are important that we are properly representing our members and the whole organisation there with governments, regulators, our membership. Regional meetings, road shows and, you know, regional IGFs and the like.
Road show address space resource request, we sent one?off for address space for the IPv6 road shows, that has been going through the ash terse who looked at the request and have agreed that that is something that we should be doing this way, so thank you very much, that is going forward. And little bits where we say, little bits of additions to member service areas that fall all under the big four sections that I mentioned earlier.
So that's basically the very high level overview without too much detail. I think I made my ten minutes. And of course yes, we also fabricate socks and T?shirts, but you know that already. They are out there, get them afterwards. Thanks.
CHAIR: Any questions for Axel? One.
AUDIENCE SPEAKER: Filiz Yilmaz, ICANN. Just a clarification of the question regarding the global policy on the the recovered IPv4 address space, that one doesn't have any reply comment period over the comments because there were no comments, so, as of the 5th April, the normal process should be moving on.
AXEL PAWLIK: We saved 21 days, thank you.
(Applause)
CHAIR: So, next one up is Andrew and then I made a small mistake on the agenda, afterwards is Niall. I posted the wrong order but I think we can live with that.
ANDREW DE LA HAYE: Good afternoon, I am Andrew de la Haye and I am going into a bit more detail on the member services and an update on some numbers and some stuff Axel has been referring to as well.
Last year, or it was the last RIPE Meeting 63 we were asked to do some numbers on ticket load from the different departments we have, for example our customer and registration services and these are two graphs on the left?hand side customer services tickets, with an increase of about 10 to 15% over the last three years. And the same is actually visible in the RS ticket load which also shows a 10 to 15% increase in ticket load. And one of the good things is that we have been processing those tickets with the same amount of people as in 2009.
Registry quality initiatives. Well we have quite some initiatives as Axel already mentioned in this area. And the key objective, again, is to have the RIPE registry complete, correct and current. The first one, the audits we performed in 2010, it was about 400. In 2011 we reached 450 mark. Now currently, we are also trying to do our utmost, although we do have a lot of requests for resources of course. And the intention is to increase the number of audits we do with an end goal of doing reviews by yearly or every second year to all our members.
More importantly, to say that we also like to provide more information to you when doing those audits in the sense that we would look at your reachability by using all kinds of tools as Atlas, for example, and that might give you some additional information on how the other side of the world is being views.
We also look at database information of course and other registry information. And we will for example will look at the certificates status you have, so whether you have ?? so that we can have a good registry in place.
Furthermore, we implemented RIPE NCC report form as Axel referred to already, with the main intention, it was a process we had for a long time already, but the intention was to improve the user experience and make it easier to report abnormalities in the registry to us and make sure that we are able to work on those.
The registry of legacy resources, just some numbers, some factual numbers. We looked at the holdings and prefixes in the RIPE NCC service region and we have about 900 prefixes in this region with 130 LIRs and we have 3400 prefixes with 2317 non LIRs. We started in February after announcing it at the last RIPE Meeting and so far, 520 prefixes were registered by 33 LIRs, and 11 prefixes by 5 non LIRs, which is interesting, because that's not the current phase we are in we didn't reach out to the non LIRs yet. But they proactively approached the RIPE NCC to be registered.
Furthermore, we got quite some positive feedback on the initiative itself. Which is good. However, there were also some concerns. There were some concerns about the clarity on the charging of how are legacy space is going to be charged and the current RIPE policy compliance is a showstopper for some. And Niall will go into some more detail after this presentation, so that we can discuss these topics as such.
Now, some more initiative 2007?01. Lots of numbers. Since 2007?01 we had 15,000, about, new registries, or new assignments under contract, which is very good. And out of the 33,000 older assignments, we now have almost half of them covered by contract and we are now in a very difficult phase where we have to reach out to those that do not respond to us.
Resource transfer listing: Well the whole listing discussion has been going on in Address Policy already but also provide a service where LIRs that have resources have a facility to show that they'd like to transfer those to another LIR. And it's all under the policy framework as such, as already been mentioned by Alex this morning.
We also tried to enhance as much as we can the member services we have. And the coordination services and one prime example was that earlier this year, or late last year, we had 128 /16, which was configured as a margin on Jew no, sir routers and that gave quite some problems to users. So we did debogonizing and we did quite some outreach and we also phoned and emailed quite some operators to explain them the situation and we had actually quite some positive responses. Some didn't like it that much that we called them but the e?mail was the preferred way. That was very clear and we also will show some analysis on the results of this, which seem to be very positive and we'll publish those in RIPE Labs later on.
Furthermore, trying to improve the way we communicate to you guys, our members, and new LIRs. For example, after three e?mails sending up and down, when dealing with a request and if we see that it's going to be difficult, we give you guys a call instead of trying to do it over e?mail, which is supposed to be much more friendly to you. And also, the calling of LIRs, our new LIR that comes in, often we see problems with the understanding of how policy fits together and what the role of registration services is. So, we also give them a call explaining them the whole situation.
We created a new tool, which is IP analyser, which will allow you to have an overview of all your resource holdings and it also shows the percentage of usage and that kind of stuff. It's a first initial setup we have, and it shows IPv4 provided aggregated, IPv6 allocated by RIR, for example, and we have a road map for 2012 to improve it even more where we will also show other address types, such as for example, legacy space if possible and overlaps and invalue it's. And later on we will also create an API so that you can use it on a local instance.
This is the picture on the left?hand side you can see all the allocations with the usage rates and on the right?hand side, you see the ranges of assignments within those.
We were also asked to do more measurements on the satisfaction of you words to us, and broader other people which are not members. So we introduce add short survey, it's like three questions or so, and implement that had in Jan to see what kind of requests you have words to us or guidance on how you'd like to be engaged. And overall, very satisfied, which is nice, but it's more important to see that it's actually being used and we do get a lot of feedback and it's very important to us to get that feedback. So we are also looking at extending it to other outward phasing departments. In the same room, we have opened life chat sessions for customer services. And well, that's actually also well received. We currently did two hours a day, but we are going to extend it as well and we are also look at extending that to different departments within the RIPE NCC.
In the training area, there is a huge amount of questions. Also coming from the 54 requests or points Axel mentioned before. And many of those were in the area of more technical help during courses and more hands I don't know experience. So, what we are currently doing is creating a small lapse environment so that people are able to do some more hands on tasks and experience us within those courses we'll give. And we are also extending the online learning and movies. One of them is the web NRs, which has been actually quite successful and we have more web NRs announced during this week and out of ten we announced, nine are almost full already. So, there is lots of interest in those web NRs. It's interesting. It's a one hour web N R on a specific topic. It's 45 minutes of explanation and there is about 15 to 20 minutes of discussion and questions there. And you can do it wherever you want. Now, furthermore, we are working on the RIPE Atlas movie. We have five new database movies and we'll show them in May. We are working on RPKI movies and webinars and v6 videos in and transition techniques.
The RIPE NCC webinars they are available to please have a look at their website and see what would be applicable for you. Currently it focuses on RIPE database and RIPE database advance topics as well as RPKI, and later this year we'll look at v6 and new LIR webinars.
There is also a little gimmick, you can make a picture in a comfortable position where you watch a webinar, and the idea is that you can upload those to our Facebook page and there is something to win, a fat boy for those that really want to be comfortable.
That's about the update. Any questions?
CHAIR: Questions? No. We are right on time. It's amazing.
(Applause)
CHAIR: Next is then Niall.
NIALL O'REILLY: Good afternoon, I am Niall O'Reilly, as many of you will know. The presentation I'm about to give has been prepared by three of us. Hank Nussbacher who can't be here and sends his apologies from Israel, Bernard Tuy from France who is in the third row there and myself. Hank and Bernhard are working in their national research and education network organisations, and I'm working in a university which is a client of our national research and education organisation.
I wanted to say a few words about services for legacy resource holders with a view to starting the balance rolling to create a policy framework, a bit late as it happens, for work that's already underway in the RIPE NCC.
So, what I want to do is to introduce and explain the situation of the legacy resource holders. To summarise just in one slide, so it won't do it justice and it will be superficial, the RIPE NCC plans, as we have understood them. I won't show any slides about other RIRs. In those four other regions, the situation is more or less the same in that all of the RIRs are faced with a certain quantity of legacy resources which have to be documented better than they're documented at the moment, and some of these registries are more advanced in their planning and implementation for dealing with the legacy resources than others. And among the four other regional registries, there are different stances, different ways of approaching the work.
The two most important slides are the ones called: "What we are looking for and what we think should be avoided" and then as usual there will be space at the end unless you talk for too long for some discussion and questions.
As you probably know, indeed as Axel and Andrew have already explained, the legacy resources are the assignments which predate the establishment of the regional Internet registries, and these resources were assigned either directly from IANA or through the Internet registry to which IANA outsourced this work at a certain stage, and even in the annual report, the shorthand expression is used: Assigned or allocated from IANA, even though in some cases, in the most of the stages of the history from been 1991 or '92 onwards, in fact IANA didn't do this detectively, but it's a shorthand which has fallen into usage Leo Vegoda has asked me to point out that it's inaccurate. And I thank him for his comments.
The legacy resource holders are early adopters of the IP technology. We come from a variety of organisations. Many of us are in the research and education community in different countries in the RIPE NCC region. And as ?? and particularly that education and research community have been sustained supporters of RIPE and the RIPE NCC and have been enthusiastic about participating in RIPE since 1989. For example, I was there at RIPE 3. Some of the other people were perhaps even there earlier. We were the member organisations in RARE at the time that RARE decided to provide an initial home for the NCC. We populated the RIPE database in the beginning with our data, which is now, of course, the focus of work that has to be done to ensure that the data has good quality, and we haven't gone away. So we are not strangers that have just come suddenly on the scene.
And here with two models of distribution of the Internet resources that. The one on the left is the one that you'll be most familiar with. It's the hierarchical model where IANA makes allocations to the regional Internet registries which, in turn, makes smaller allocations to the LIRs, who, in turn, make assignments to their customers. The legacy resource holders didn't have RIRs there to mediate the allocation, the assignment of resources to them in the day and indeed, as I have said, IANA didn't in the later stages of the history, from about 1991 or 2, IANA was working through a global Internet registry run at various stages either by SRI or by network solutions. But the point is that we have two streams of allocation and assignment of resources here. The one that we are most familiar with and that has given ?? that has been mediated, supported, implemented by the regional Internet registries, of which RIPE NCC was of course the first, and the other stream, which is no longer flowing, in that the legacy resource holders can't just because they are legacy resource holders, go back to IANA and look for more, and we don't maintain the registration of those resources with IANA. We maintain them with the regional Internet registry for our region. And the crucial thing is here we have obtained the resources that we hold, not through the distribution policies of the RIR, but through a historical policies and those are the basis for our resource holdings and our relationship with the regional Internet registries is that we need, we provide information for the registries in their database, and we need access to that information that's in the database. And so, the significant part of the relationship between the legacy resource holders and the regional Internet registries is database access.
Now, the situation we have today is that over the 20 years that we're celebrating at the moment, the RIPE NCC and the legacy resource holders have drifted apart. And this has been because the NCC has had to focus on the supply chain of new Internet resources, and by and large, although that's something of a simplification, the legacy resource holders haven't needed to be part of that distribution chain, because they got big allocations in the very early days, and sometimes they have needed to augment those with conventionally distributed resources, but by and large, their legacy holdings haven't really been a matter for discussion with the NCC.
So, you can see this as a kind of mutual neglect over the years because it wasn't imperative from the business reasons, either of the holders or of the NCC on the other side. And it's not such a mutual neglect that we could talk of a marriage gone silent. We are not at that crisis point, we ?? it's time to, as I say at the bottom of that slide, we need tone gauge and to work together. And a consequence of this drifting apart has been that the data quality, the quality of the data that represents the legacy resources has degraded over the years. People have moved on to other jobs who are still listed as contact persons. Telephone number systems have been renumbered so the contact phone numbers might be different. People have rebranded their domain name, so the e?mail addresses might be wrong. On the other hand, quite a few of these have been kept up to date, and I know that I have submitted a correction to the one which represents the /16 resource that my university holds since the Vienna meeting.
And I tried ?? the changed log on that particular resource is very interesting. It goes back to the earliest change entry I think in 1990 is stamped with the e?mail address dfk [at] cwi [dot] nl, which some of you will recognise. But the thing is now that the NCC is beginning to focus much or actively on registration and on added value service based on that registration, it's very important that we engage and that we work together and that the contact that we now re?establish has the right basis and is done without any unfortunate misunderstandings.
What I understand, and this is on the basis of what mainly on what I have heard at Vienna, because these slides were prepared mostly before coming here, but I haven't seen any major contradictions. And in fact I know that before I ?? before I came to stand here, the NCC staff have taken a look at it and I haven't been advised of any major diplomatic mistakes on my part, so this is more or less accurate.
The proposals from the NCC for dealing with legacy resource holders were announced at Vienna by Axel Pawlik. There is a corner of the NCC website which explains, on a number of pages, what's about to happen, how the approach is and essentially there is a four stage outreach plan. You have heard already of that from Andrew. The options for the legacy resource holders are to become members or to register their resources through a sponsoring LIR. More or less along the model that we have had before for PI. It's proposed that the holdings will immediately become, on registration, subject to all RIPE policies. The motivation that has been explained to me for this plan is, so that the RIPE NCC can act as a responsible registry, and significantly, this activity isn't based specifically on a policy. The situation of the legacy resource holders has also been considered by the charging taskforce and in their report there is a reference to a carrot and stick approach to the legacy resource holders. And that's more or less it.
If there ares were details, I am sure that Axel or Andrew will contribute them in the discussion.
So, what we're looking for as legacy resource holders is that there should be a clear policy basis for the activity that's already underway. We know that policies don't come out of thin air by magic and some of us will have to do some work and I expect that I'll probably be one of those. I can't ?? my authors haven't agreed to let me commit their time and effort for it but I am sure I won't be on my own. But it's really important that the policy that we arrive at is derived from consensus of the stakeholders, and this is a little different, the concept of consensus here is significantly different from what we usually consider to be a consensus in the RIPE community because we are talking about a policy and plans based on that policy which are particularly directed words to a sub constituency of the RIPE community, and that sub constituency institutes the primary stakeholder group of this policy and of the activities used to implement that policy. So it will be important that it won't be the rest of the RIPE community who decide who should be the policy that targets the legacy resource holders. It needs to be a consensus which includes the legacy resource holders.
Apart from policy, we also need to have clarity as to charges and corresponding services. Now, the charging scheme, I know, is a thorny problem and it's quite complex already, even without the legacy resource holders. But it really is important both in terms of offering membership to some of the legacy resource holders and of having local Internet registries sponsor the registration activity is very clear that nobody has any ugly surprises with regard to charging, and I think that's worked for the study, in particular I am not supposing that by standing here today I am going to present any solutions. This is really a suggestion that we are going to have to do some work and hopefully we'll have significant programme done before Amsterdam.
Now, some of the degradation of the data quality that has happened over the years has been awkward for the holders of the resources because data that they have chosen to use the RIPE database to publish has somehow become damaged and whether this was through smaller RIRs in the bulk update process that happened ten or so years ago in the great ERX project, or whatever it is, there are operational needs to change some of the data very urgently, and I have heard from some of the people that they have been more or less coerced into signing up to the legacy resource service agreement in order to have those operational considerations addressed. And it's rather important ?? it's really important that, from our point of view, that if there are such operations imperatives to get the data corrected, that that should be ?? those corrections should be done before any freeze of the data that might be considered.
What we think should be avoided and you may be surprised at how harsh the first piece of terminology is, but this is actually ?? this isn't a term that I have invented, it's a term that has been fed to me by one of the LIRs on behalf of some of their clients. The RIR responsibility approach that the NCC has used, has been seen as heavy handed, and this is something that we need to avoid, and as I mentioned already ?? no, I didn't say that already. I mentioned already that the RIPE NCC was proposing, in the current activity, that on registration, the legacy resources should be submitted to all of the RIPE policies. I think that's too glib. I think that needs to be much more carefully examined to see which of the policies are actually useful and which of the policies are applicable to the legacy resources, because otherwise, the resource holders, if they are doing any kind of due diligence, will send this to their legal department and it will stall there for a long time and that will delay the rectification of the data and be counterproductive for the goals that the RIPE NCC and the legacy resource holders alike, share.
And whatever about the carrot and stick approach suggested in the review of the charging taskforce, if the carrot is out of reach, and it might only be people from Amsterdam will understand this, there might be a green grossers shop just over there, but if there is a deep piece of water in the way and the next bridge is a kilometre away, that greengrocer is a long way away. So if the carrot is out of reach. All we can feel is the stick. Please don't put us in that situation.
CHAIR: Thank you. Questions?
AUDIENCE SPEAKER: Shane Kerr from ISC. I am not a legacy resource holder or an RIR, or LIR, so maybe it's a bit weird for me to talk here. I am curious why this discussion is happening here and Address Policy Working Group.
NIALL O'REILLY: I thought carefully about that and I considered that there is no Address Policy involved here. The legacy resource holders hold their resources through a historical process which is no longer in effect, and that process is not reachable by Address Policy today. But rather what's needed is a clarity about the services that will be offered by the NCC to those resource holders. That was my thinking. If it turns ?? if we agree, I am quite happy to do this ?? do the policy development process in the ENUM Working Group if people think that's the right place to do it. I am not weathered to doing it here.
CHAIR: I think Keith was next.
AUDIENCE SPEAKER: Keith Mitchell. I am making this comment in my capacity as a member of the arbiter panel. You mentioned that there was some data scrambling as a result of the RX process. Just to confirm that there was one arbitration case that I did where old bad data got over written ?? old good data got over written by ancient bad data. So yeah, there is definitely stuff in there like that. Rightly I'll thanks for confirming.
AUDIENCE SPEAKER: Gert Döring, Address Policy Chair, speaking with my hat on. I tend to agree with Niall, not only because it's saving me work but also, well Address Policy sort of handles addresses given out by the RIPE NCC and maybe returned to the NCC and traded around in the LIR framework and not stuff that's also stuff. So, I think services is fine.
AUDIENCE SPEAKER: I am not wearing the hat of the member of the charging scheme taskforce right now. I am the member of the academic community from Poland, also an LIR and a holder of the legacy resources. One comment ?? sorry, one question and a few comments.
The question is: In the name of what person of legacy space holders are you speaking? Because you said many times "we" that's a question, because it's very important. It's about your situation or about the situation of thousands of legacy holders? Will you discuss that on the mailing list, whatever.
NIALL O'REILLY: We have a mailing list. We have a small engagement in that mailing list and I am anxious to have that engagement widened and when I have been saying "we" just now speaking, I am speaking not only for my authors but for a relatively modest number of people interested either as likely sponsoring LIRs or directly as legacy resource holders whom I have been able to talk to. And I am anxious to make that number wider, and I am very conscious of the paradoxical position, if you like, that I put myself in when I say that it is important that any consensus that we achieve is a consensus which involves the primary stakeholders because, just as the body of LIRs cannot be said to be representative of the legacy stakeholders, neither can a small rump of legacy stakeholders, so, I am really anxious, and this occasion is especially a flag waving occasion, you know, if your interested in being part of forming that policy because you are part of the primary stakeholder group, then please contact me, please join the discussions, please be part of developing that policy because the fewer people we have there, the more the policy is open to challenge, and the less the policy can be useful either for those of us who contribute to its development or for the RIPE NCC. And I am really glad and grateful that you have asked that question.
AUDIENCE SPEAKER: The more PR we have, the better.
NIALL O'REILLY: Absolutely.
AUDIENCE SPEAKER: That was the question. And two comments ?? two or three whatever. The first one is about the situation that you were trying to update the record. I am both a legacy resource holder and a LIR, but most of our friendly universities, they are not LIRs, and both we and our friendly universities have no problems with updating the database records.
NIALL O'REILLY: This was another resource holder, not me.
AUDIENCE SPEAKER: Okay. Maybe it's a kind of bad luck or I am lucky enough to ??
NIALL O'REILLY: On the other hand, it was more than one and it sounds as if there is a problem ?? a systemic problem that needs management attention. I don't say a big problem but one that needs attention.
AUDIENCE SPEAKER: Another comment is about this responsible RIR, I am pretty much sure and probably Wilifred will confirm that that RIPE NCC, and other RIRs, are enforced by the contract with ICANN or IANA to have the database filled with proper data, so it's not about being irresponsible, it's about if you go filling the contract. So, it is kind of policy. They have a contract in name of us, of the stockholders, or am I wrong?
NIALL O'REILLY: If Wilifred wants to talk next, I think...
WILIFRIED WOEBER: I am wearing two hats at the same time. One has being Vienna University and as such, an organisation which holds such an early registration. Second hat: Very close to that is a long time established LIR, and the third hat is ICANN activity and member of the WHOIS policy review team.
Regarding the contract, it was pretty interesting for us on the WHOIS policy team to finety find out how narrow and extremely narrow, actually, the mandate and the capabilities and the contractual stick of ICANN is in that environment, because I think what you are referring to is the registrar accreditation agreement, and this only applies to a very small number of organisations being active in the domain name business, so this is far from our numbers business here. Of course there is a, sort of a high level, I think, general understanding in the community that we want to have complete and correct data in each and every registry, be that names registries or be that numbers registries, but as this community, I think we should be very careful to mix the numbers problems with the names problems. And I would suggest that we, of course, have the existence of ICANN in the backs of our minds, but sort of not to take that as the primary controlling environment, but to look at what our community needs to discuss and needs to resolve. Sort of just on that one.
And then I go back to the back of the queue because I have got a couple of other comments and there is other people at the microphone.
CHAIR: I think it was the back microphone then.
AUDIENCE SPEAKER: Hi, IPv4 market group. You had the point that the legacy resource holders may be subject to a bunch of policies. Have you looked into which policies yet and specifically, are you looking at the piece related to verification of needs and would that apply to legacy holders?
NIALL O'REILLY: I am not sure to whom that question is directed. You'd like me to explain what I mean in the first place?
AUDIENCE SPEAKER: Well, which part of policies would apply to ??
NIALL O'REILLY: Well, the situation at the moment is the template registration services agreement from the RIPE NCC has a very short one sentence saying that, regarding policies which says that on registration, the legacy resources will become subject to all RIPE policies. I paraphrase because I am not ?? I haven't learnt the exact text off by heart. So the difficulty for ?? the difficulty that I imagine, for example, to make a very concrete case, that the legal department of my university would have in dealing with that, is that it doesn't numerate the policies. It imposes homework to be done on whoever is tasked with the due diligence, and it will cause a delay in dealing with the case. And we saw the display outside, we saw that there are 4,000 or so registration transactions to be done, and if any of them are systematically delayed through having to do and to repeat due diligence, that could be made unnecessary by itemising the applicable policies and discarding from that list the ones that were found to be inapplicable, then the process would be much simpler for bringing these legacy resource holders and their resources into the registry as is desired. Now, you asked specifically the question about needs analysis ??
AUDIENCE SPEAKER: Here is my question then: So, if, after the legacy holder is updated in the registry, say subsequent to that the legacy addresses are sold in a transaction, will those legacy resources be subject to all the policies within the RIPE manual?
NIALL O'REILLY: That's a downstream process and out of of scope for what I am addressing here, but Kurtis is nodding.
AUDIENCE SPEAKER: What you said about applicable RIPE policies doesn't apply to downstream transactions or downstream occurrences?
NIALL O'REILLY: The urgent task to be addressed is updating the data for these resources and that's a thing that should not be stalled by any administrative stance that is chosen. If there are, then, downstream transactions that are governed by policies around those transactions, I expect those policies will apply. I don't know. Because we are here ?? I am here kicking off a policy development process. I don't know what the outcome of that process will be, so, I can't guess whether there may not ?? I'd be very surprised if there were special exemption clauses for the kind of thing that you need but I think that's speculative and premature and from Kurtis's body language probably not the best use of time at the moment.
KURT LINDQVIST: If you sign as a legacy resource holder you sign the contract right now as it stands, then your address space will be subject to all existing and Nick policies. There is no difference between that address space and that being awarded by an LIR. Whatever policies exist, that ceases to be legacy that point.
AUDIENCE SPEAKER: Okay. That may raise, if I can just comment quickly. That may raise an issue for legacy holders insofar as when there is a transaction down the road then, say there is a transaction within the RIPE region, a buyer and a seller, so the seller is a legacy holder, the buyer is someone in the region and the buyer then has to go through justification of needs.
KURT LINDQVIST: Correct. I don't think we are going to make policy assuming everybody is going to sell their address space, but yes.
AUDIENCE SPEAKER: Nigel Titley, Chairman of the board of the RIPE NCC. Just a very quick note. The charging scheme task force has delivered its report. The board is due to consider that report starting in June of this year. Could I urge you to at least give us some idea of what your policy might say fairly early on so that we can consider that?
NIALL O'REILLY: Noted.
WILIFRIED WOEBER: I don't know how much time we have got left so please feel free to cut me off. One observation to start with. I was pretty much surprised when, within the framework of the Vienna meeting, the whole thing was sprung upon this upset of my constituency out of the blue without any previous discussion on any of the mailing list or without any alert. Just as an observation. Personally, I was warning my constituency for about one?and?a?half or two years that in the aftermath of 2007?01, something will probably be happening to the old address space, but still procedure wise, I was pretty surprised and I was slightly disappointed at that point in time. Just for the record.
Secondly, there is, to a very large degree, I agree and I support your lines of thought regarding the issues of completeness and correctness of the data and what the procedural ramifications, or preconditions are to continue being able to use the registration services. But, just for the full community here to which I am in and to start thinking about it copying what we had during a private discussion yesterday in the afternoon, I was relating the current situation we are having in the numbering environment with a situation that ?? if I remember correctly ?? Belgium had a while ago, in a completely different environment and that was the lack of drivers licences, so, historically you just bought a car, you sat in the driver's seat and off you went. And then at a certain point in time, the community said, okay, we will align with other states and nations around us and we'll require the drivers to hold a licence. This is, maybe, pretty farfetched, but I think there is some similarity to this community here, because we are seeing a development in the Internet of some 30, 40 years, and some of our constituents are holding address space since 80, whatever. The world has moved on and I think there is quite a good reason to suggest that, sort of, those old kids on the block should eventually be subject to the same regulations or the same processes or the same procedures like the people or organisations being born later on. That's not saying that well we should just sort of smack them on the head. It's just sort of an idea to think about, that eventually you should align the old stuff with the current stuff. But the procedure should be well developed. Thank you.
KURT LINDQVIST: I'm really going to cut the microphone off, that's a shame ?? you have got half a minute each, max. Shane. You get the last half minute.
SHANE KERR: So, it wasn't clear to me who you expect to be making this, the policy ??
KURT LINDQVIST: I was going to come to that. So, I wanted to say two things. One to the presentation, you kept talking about the primary stakeholders and I'd like to remind you that, as far as I am concerned, as a non?Working Group Chair, we are all stakeholders in the address policy and the address usage, although there is some that have a right to them, we are all affected how these addresses are used, which I guess is why you came here.
But for the process, if we all seem to agree, all ?? the only one who said something I was Gert, that this is not belonging address policy, if Niall and the legacy holders come up with a proposal policy or proposal how to handle this, do we believe they should right is up as a policy and take it to the NCC Working Group. Is that okay with you?
NIALL O'REILLY: I think it's important that you put that question to the meeting. Yes, please.
KURT LINDQVIST: That's your intention?
NIALL O'REILLY: That's the plan at the moment.
KURT LINDQVIST: Is there any objection to doing it that way? So all in favour? That's kind of a few hands. Anyone against? Okay. Well, I just want to make ?? I am not talking about a policy itself here, I am talking about a process for looking at a policy. Okay. We'll do that. You write it up. We'll take it to the NCC Services and we'll work on it from there the thank you.
(Applause)
KURT LINDQVIST: Next is then Andrea. Now we are behind schedule.
ANDREA CIMA: I'll keep it quite short, as we are running out of time. Good afternoon everyone, my name is Andrea Cima, a part of the registration services at the RIPE NCC. During the last NCC service Working Group at the last RIPE meeting in Amsterdam, I gave a presentation about what the RIPE NCC was doing, how well we were getting prepared for the moment in time before reaching the last /8. And today I would like to give you a brief update about that.
The current situation, we are all aware about it. IANA deleted its pools of IPv4 address space in February 2011. APNIC reached the last /8 after and we expect the RIPE NCC to reach the last /8 any moment soon.
Now, if we look at the IPv4 distribution, these are the number of allocations made by the RIPE NCC over time. We see that there is a generic growth trend. However if we look at 2010 and specially 2011, we see that this growth is slightly steep err, especially 2011, where we made a record allocation number. Now, keeping this number in mind. If we look at the amount of address space actually issued by the RIPE NCC, if we look at 2011 at the amount of address space issued is in the same line of 2008, 2009, more or less. So, what does this mean? It means that we have issued a higher number of allocations, with more or less the same address space.
Now, we can see that this is looking at the numbers, we can think this is as a result of the policy Run?Out Fairly, which is a policy which was accepted by you, by the RIPE community in December, 2009, and this policy gradually reduced the allocation and assignment periods. And these periods are currently set at three months. Now, it was really clearly stated also by the authors in the policy proposal that this proposal did not intend to stretch the lifetime of IPv4. But the goal was to ensure that location and assignment of v4 resources were being done in a fair way. Meaning, to avoid that one organisation would come, empty the pool and have address space for another year or two years while the organisation next in line would receive nothing at all. If we look at the numbers. We can see this this policy has been effective and has reached, achieved its goal.
This is the RIPE NCC IPv4 pool. What we still have available nowadays. As you can see this is decreasing steadily. But there are no big jumps in there. There are no drastic changes. And this also shows that so far, the RIPE NCC membership has ?? there has been no panic in the RIPE NCC membership. And this chart can also be found online on our website.
Now, moving to the processes. What I want to say is that we did not change anything in the way that we evaluate resource requests. We did not change anything in this because the evaluation processes are based on policies which are set by you, and if there is no policy change, also the way that we evaluate the requests will not change. However, what we have been doing is fine tuning our business processes to further increase due diligence, fairness, consistency and transparency.
We have also divided the moments, the period before reaching the last /8 into three phases. The current phase, phase 1 in which we are now. Phase 0, sorry, Phase 1 will start when we have one month left of address space or we reach the last /10 available and we will enter Phase Two when we reach the last /8. For Phase 0, it's a current phase in which we are now. The main change that we have implemented so far is how we handle the tickets.
All IPv4 tickets are now redirected to a separate IPv4 ticket queue. And the main change as well here is that while, until seven months ago more or less, you were used to dealing with the same IP resource analyst from the start to the end of your request, this is not the case any more. What happens now is that you send a request, the first available IP resource analyst will take the ticket out of the queue, send your reply in case more information is needed. When you reply to that e?mail with the additional information, the ticket will not go back to the same IP resource analyst but will go to the bottom of the queue and once it's at the top of the queue the next available IP resource analyst will take that ticket and continue the evaluation. And it is done to enhance the first come and first serve principle.
Furthermore we have a lot, lot more information on our website. The first one is the one I showed awe few moments ago with the amount of address space available. Furthermore we have added a graph which contains the number of hours that are requested for IPv4 address space is waiting for a reply. And also a graph that shows the number of IPv4 requests still sitting in the wait queue.
Furthermore, we are currently informing all requesters for IPv4 address space and organisation in the process of becoming an LIR about what the current situation is avoid possible surprises.
As previously mentioned, we will reach Phase 1 when the RIPE NCC has one month or a /10 left of address space. Now, at this moment there are some additional changes that will occur, and the first one is that APRAs will work in pairs, so physically two IPRAs will be sitting behind one computer and look at your request together. And the main point of this is to increase consistency. We know that this may lead to longer response times. We know you don't like that. So we will do our best to keep the response time as short as possible. However, due diligence at this moment in time is the highest priority.
There is also a change in the call back procedure. At the moment if you call us, we will call you back within two hours. However, when we reach Phase 1, we will call you back only when your ticket is at the top of the queue. We do this to avoid that someone that calls gets higher priority than the other tickets in the queue.
Another change is that tickets that can approved, resources that can be issued will not be approved right away during the day. But they will be moved to a separate approvable ticket queue. The approvals are delayed until nine o'clock the next morning, so that we have additional time to check for consistency and additional time to make sure that all the allocations and assignments will be made in the correct order.
This means that the requests for IPv4 will be evaluated by IP resource analyses between 9 and 7 and its collations of larger requests will be dealt with between five o'clock in the afternoon and nine o'clock in the morning by RIPE NCC management. And this happens with large resource requests, they are usually already evaluated by two IP resource analysts and reviewed by RIPE NCC management.
Another important point is that if Phase 2 ends on a Friday, or the day before a bank holiday, we will enter Phase 2 the day after. So we will do that on the Saturday or on the bank holiday to avoid speculations.
Phase 2, we will enter Phase 2 when we reach the last /8. At this point in time members can only receive a /22 of IPv4 address space even though they can justify a higher amount. Another important point is that the LIR must have an IPv6 allocation, either from the RIPE NCC or from another LIR. And from this point, according to current policy, no new PI, provider independent assignments will be made for IPv4.
Here is some links, also to the graphs I have previously mentioned. And I don't know if you have any questions. Before, I also want to say something. Tomorrow at 3:30 we have a questions and answers session at the info hub about reaching the last /8, so if you have any questions you can also come tomorrow and we can have an informal discussion about that.
KURT LINDQVIST: Thank you. Any questions on the last /8? That seems a lot more clear than legacy space. Thank you.
(Applause)
JOCHEM DE RUIG: Good afternoon. I just want to update you on the order we have received from the police in November last year, where we are now, and what further steps we're taking.
So just to bring you up to speed. We received an order. It came across from the Atlantic regarding DNS change err case, and the F B I requested the Dutch police, via procedure they have called the EMLAT procedure, the mutual legal assistance treaty process to free some blocks in the RIPE database.
So, we received that order based on a specific article of the police act, and after that, we executed the order, because it was quite a time, a tight time frame they were working about. And we also started investigating the legality and the obligatory nature of the order which they told us was compulsory. So we looked into that, and actually, the Article 2 is quite broad, and the Dutch Supreme Court has made a ruling about it, that the article can order people to tolerate a situation, but it can't actively enforce something, which they were actually asking us to do. So ?? well, they made some more statements about it. So what we did then, we went back to the public prosecutor and we said okay, can you explain us the legal basis of the order, why you feel that the RIPE NCC has to execute the order? Well we didn't really get further legal basis. However, we did get ?? well, I wouldn't say a threat, but he did say if the order was reversed that he might consider confiscation. He was not very clear about what that would be, but he was talking about RIPE NCC administration. I don't think it was very specific.
Well, we had some more communication and in the end he confirmed he wouldn't proceed with the confiscation of any administration from us, so that meant that, that implied that we didn't have to follow the order any more, so we actually reversed the order in January, so we sort of unfroze the blocks, and we informed the relevant members, as we had done in the beginning of November as well, about the situation.
Also, we wanted to get more clarity on the situation and to get a ruling, so, we actually charged the state and we summoned them and we are in the process of legal proceedings. And just to give you a bit of idea where we are at the moment.
We had our first hearing on the 14th March, it was scheduled. The State didn't appear at that time, but they did appear on the 11th April, they have this window, and now they are writing up their defence in about six weeks, and maybe they can get a small extension then and after that, the Court can choose which way to go; whether they do a hearing or do a second round of written statements and then probably somewhere in 2013, hopefully the beginning, we'll actually get a ruling on this.
Furthermore, we also started looking at how do we want to deal with this in the future? We took a step back and reevaluated the procedure and what actions we took. And we wanted to go back a bit to basics. So what we looked at is some crucial things we feel we have to comply with when we deal with orders. So, all activities are for the benefit of the RIPE NCC membership. Well, an accurate registry is crucial for that. And accurate data is to record the organisation person responsible for those IPs and changing or amend that go data is actually not beneficial words to that membership.
So what we derived from that is that in the future, we want to follow just Dutch court orders, so not an order based on a police act or any other way, and we'll look at these orders of course on a case by case basis, and check the legal basis and whether the procedures have been followed correctly.
The action that is asked from us we'll check whether it's appropriate and proportional. So what we mean by that is we want to make sure that the membership is not too much affected by this order, otherwise we will not follow the order and we'll just challenge the order. And we would challenge an order when it jeopardises the accuracy. So if people purposefully change data in the public registry.
The last point I wanted to make, I think this was quite a long process and it was quite some communication. The big issue we had I think on the side of the Netherlands there was not like extensive communication between the RIPE NCC and the public prosecutor on the content of the order and actually that's why we're partly in this situation.
So, to be continued...
We hope the proceedings go as quickly as they can, and then we'll report back. We'll try to be as transparent as possible so when there is an update, we'll publish that on the website. We have published the summons and we published several news items on this. So that's it from my side, Kurtis.
KURT LINDQVIST: Questions?
RUEDIGER VOLK: When this happened I was on travel and didn't pick up a lot of information. When I came back, I figured out that there was actually some recognised customers involved and then I looked back at the communications that the RIPE NCC put out, and I figured out well, okay, you didn't really identify what was going on. While, on the other hand, the FBI and so on had done extensive press releases. Was there any consideration on your side to actually tell what was going on? Maybe just giving pointers to press releases that are relevant?
ANDREA CIMA: We did inform the members immediately when we got those requests, and we informed them about where this was coming from. I don't know if they asked about details, what this case was about. I don't think so. But, I would have to verify.
RUDIGER VOLK: When you are saying the networks, I think you mean the resource holders, not the networks that were connected and would have customers that were using the resources? Those networks might have been actually benefitted from knowing what's going on. Similar okay. So, we could improve on that.
KURT LINDQVIST: Okay. Any other comments, questions? If not, thank you.
(Applause.)
And then last up is Daniel.
DANIEL KARRENBERG: So, this is the services Working Group and I'm actually talking about, not about all the successes but about shutting down some services. But before I am going there, I want to tell you what the strategy is behind that and I'll do it rather quickly. There will be an extensive opportunity to discuss this in the measurement analysis and tools Working Group tomorrow at four o'clock, so this is just a warning because we are talking about NCC Services.
So, this is what we got, or part of what we got from the membership survey. And these are actually quotes about our information services, where can you find things? And so on. And it was quite strong, I would say. It was the area, I think, where there was the most discontent. And it wasn't so much about people thought the services that we were having there like RIS and so on, and TTM were bad, but it's difficult to find and difficult to use. And that's something we want to do something about.
One of the reasons why this is happening is that there is so many of them, and we are not really quite consistent in giving awe good road map there. There is just too many. So, where we are going, and that's what Axel has already alluded to, is to get these information services and where you actually get information and can manipulate it into sort of those roughly four areas that are here. RIPE Labs, we all know that; RIPE Atlas, the active measurements; RIPE Stat, everything you ever wanted to know about Internet number resources and the LIR portal, very important if you are a member to administer your resources and your NCC membership and Andrew has just told you about IP analyser that is part of the LIR portal to make it easier to manage your resources.
I'm going to concentrate on the first three, because they are open to the general public, whereas the portal is just for members.
So, what are these? RIPE Labs is about stories, and we are actually quite good at that, it's more or less a blog, a similar format. You read it a lot. It's quite successful. And it's a good platform for us to bring stories to you to gather feedback and so on.
What we want to add to RIPE Labs in a more consistent way, we have it already but in a bit more consistent way is general statistics. We have, for instance, what we published in RIPE Labs the IPv6 RIPEness, somebody just recently an article about the membership demographics, you know, what happened to the membership in the booms and busts and so on. But we want to do that a little bit more consistently so that you can actually, if you are looking for a specific statistics, they are easier to find.
RIPE Stat is about specific IPs, IPv4 and IPv6 and ASNs. If you want to know something about resource, what we want you to do is go to RIPE Stat. There will be everything there. The registry information, you know, RIPE database and so on, the result of active measurements, results of RIS, other data, you know, if you type in an IP and you get to know everything about it. And that's actually come quite a long way, and I think it's actually really nice right now already, but we are further improving that.
And then finally there is RIPE Atlas, that's about the state of the net, active measurements, but it has two components: One is the big picture, so the Internet traffic map, if you wish; and the other is the individual measurements that you can do, we call it user?defined measurements. This is the system we want to get to, and the categorisation we want to get to. And of course there are links between threes three obviously.
The other thing that we heard quite clearly from the membership survey is that you are looking for additional benefits, or exclusive benefits to the membership. What does it give you other than the registration to be a RIPE NCC member? So what we want to do in all these three areas is ?? while obviously it's in our all our interests to have a lot of this information public and available to the community at large, but even beyond the community, for instance, to the law enforcement community, to policy makers and so on, journalists even, we are going to develop some parts of this that will be available to members only. And that may be in a number categories, one would be information that's only available to members, but another would also be that kind of special analysis and value added analysis and things like that would be available to members only. And we already are moving a little bit in that direction in RIPE Stat with recent things on the object browser and the hierarchy browser. We also have that in RIPE Atlas where we open up the user?defined measurements to the membership, that will happen tomorrow.
So, this is the strategy. But there is a bit of a fallout here because we have some service that is we all know and love. One is TTM. And the immediate question of course that I got when we were developing this strategy is, you know, what about TTM? Well, my read about the TTM service is that it's from this era and for those of you who don't recognise this, this is TTM multiplex err gear, where you are basically, the bandwidth scarce times. It's incredibly sensitive measurements that were intended to help with capacity planning in a time where transmission resources were very, very scarce compared to now. But really at the moment we are here and the one way delay and all these things are not really all that interesting any more, and we got that also quite clearly when we asked the current TTM customers for feedback. So, it's not really something that's, you know, current any more.
What's more, what I found out when we were sort of analysing this is that actually the equipment that's out there, including the central infrastructure that actually shows you all the graphs and so on, is really, really ageing and we have really failed there, I have to admit that, at having a life cycle replacement programme there, and keeping the stuff up to date. So, that actually is a major challenge.
So, we went and in Vienna, we had a workshop with the TTM users, we have talked to many of them in the meantime and what came out of it is, you wanted to know what is it in TTM that people are interested in and we found, not to our surprise, that the time service provided by TTM boxes since they have a time clock, is very, very popular. Everybody wanted that. But that's not really our role as the NCC to provide, you know, time service. Well maybe it might be, if people want that. There is also quite clearly some people, the crowd that has, the people that have TTM boxes are generally interested in measurements, and to know a bit more, have a bit more hard data about where they are in the network and things like that. So that was also clear, people said they want measurements. But what they also said is like we want some sort of not compatible measurements in more than one sense. Our knock staff is not really familiar with the way it's presented currently but the TTM results are presented, that's what we heard, we heard that one way delay measurements are not really what normal network engineers are used to. Round trip stuff is what they understand. And also there should be better alarms and things like that.
So, we looked at the options of either sort of totally revamping the TTM thing including deploying new hardware worldwide and getting the time sources up to speed again, and revamping the whole internal workings of it. And the other option, to actually look at it and to tune it down a little bit, remove the requirement for GPS clocks and integrating it into Atlas. Because, one of the things that we were planning for Atlas anyway, is to have what we now call Atlas anchors, those are Atlas boxes that are both very powerful probes, so they are not the tiny things, but they are more server like, but also possible destinations or willing destinations for Atlas measurements, where, and that's one of the ideas that Geoff contributed was to actually say, oh if the probes probe to an anchor point, then at the anchor point we can also measure on this measurement traffic and do something with it, and that's one of the ideas. And it's also a way of getting some level of distinguishing the two ways of round trip but not with the GPS accuracy.
So, based on that, we decided to, and that's the RIPE NCC, decided to shut down the TTM by the end of Q 32012, and I know there are some concerns about the process here, and we can address them. But what I really want to discuss here, and in the mat Working Group is the decision itself, not the process. We can talk about the process too but let's not mix the two.
We announced that a couple of weeks ago and we told the TTM hosts we will give you a best effort help to continue the time service. In practice that would mean you get the route password to the machines and we help you with the NTP configuration if you need any help, it's not all that difficult. We will also, we know that some TTM hosts use it for more or less internal measurements, and also some SLA monitoring is involved there, so we'll help you to localise this, and our friends at NIC PR actually have a software that might help with that so we will help with that but it will be best effort and we also have a stack of clock cards still lying around, so if one of them burns or one of the GPS receivers, we also have a few of them left, burns, we will ship you one. But, only on a best effort basis.
At the same time, we are going to develop the RIPE Atlas anchors and I am not going to go into those details again. And actually to, it was base ?? because this was based on the feedback, it wasn't that surprising that the reactions of most of the TTM hosts were quite positive about this.
And the further work to make the detailed plans or adjust those plans, that's for the mat Working Group.
Then of course, it's all a domino, right. What about DNSMON? DNSMON is the DNS monitoring that we are using and that we are providing which has been quite popular recently when there were some threats to the DNS infrastructure. But, unfortunately it's based on a TTM boxes, so we have to find and alternative an DNSMON is still very much used. So, head scratching again. But the good news is that actually RIPE Atlas can do the same measurements today. It's not reviewed in UDM yet, but virtually all the measurements that the DNSMON measurement infrastructure does, can be done by Atlas. So that actually gets better because you have thousands of Atlas probes, they are better than sort of tense of TTM boxes, but Atlas probes are not TTM boxes so there is a bit of work needed. Also thinking about how to use the Atlas probes, there is also a question, what we got from the feedback is that a generic alarm service would be good and that's also what we get from the DNSMON customers. You know, we don't only want to have the graphs but we also want to be notified when, you know, something turns red. And also the generic presentation and visualisation would be good.
So, we think we can move the DNSMON to Atlas in time before the Atlas probes go away, and we can do the generic alarms and presentation through the way we are going to do it with Atlas, which is actually within RIPE Stat. That work's also being done in the MAT Working Group.
So, if you are interested in the details on how to make these two things work Andy involve these services in a positive way, the MAT Working Group, Thursday at 4 o'clock in the place.
And that brings me to the end. Any questions?
AUDIENCE SPEAKER: Hi. Eoin Mikel. Can I just point that we won't have a great deal of time for discussion about this so I'd like to encourage people to use the mailing list as well in advance of the session.
DANIEL KARRENBERG: I didn't get that acoustically.
KURT LINDQVIST: We don't have very much time in MAT so we want to use the mailing list.
DANIEL KARRENBERG: Whatever. But if the MAT Working Group is the forum. We also have an Atlas BoF coming up, so if people want to discuss the specifics to DNSMON, Atlas and things like that, that can be done in the Atlas BoF.
AUDIENCE SPEAKER: Brian Nesbitt. I know you said you didn't want to discuss the process. You wanted to discuss the decision, but as I have practically no problems with the ultimate decision, I am going to discuss the process.
In Vienna we all sat down and we talked about this and the next thing I heard as a TTM operator was this service was being shut down without as far as I was aware, any ultimate input from the community or any policy or any notification on the MAT mailing list or anything. The next notice was we're done now. And while, again, as we have discussed yesterday, I'm not going to fight against the close down of of TTM, I think that there is a lot of good things in Atlas, and it's probably ultimately the right decision for measurement and the NCC Services, the fact that this was done in the way that it was done is unacceptable. And I think that this is the forum. The other stuff we can discuss in MAT, but this is the forum to can you say that.
DANIEL KARRENBERG: I wasn't hiding from you, I was standing next to the monitor box to be able to understand. I think unacceptable is a very strong word that you are using here. And I take offence, actually to the word "unacceptable." I agree, absolutely, that the problem could have been better. All I can say is that the intention was to actually have this discussion in the MAT Working Group starting in January and actually that's what was said in Vienna, we would come up with a plan in January. And what actually transpired is that for a number of reasons, which are totally within my responsibility and some of them are force majeure, we didn't do that, and I woke up in March and basically said, hell, nothing happened but we really cannot let this simmer on a little bit more. With hindsight I probably should have said in March, okay, we lost the time, let's go to Ljubljana and discuss it in Ljubljana, but my intention there was to actually move forward, and with hindsight, yes, we should have done a little bit ?? we should have done it as we planned. We didn't. I'll take the responsibility for that. But the important thing is here is that actually there is, and I still believe that, consensus on the way we move forward. Not just the problems could have been better, but "unacceptable" is a bit strong, in my opinion.
BRIAN NISBET: The reason that I use the word "unacceptable" and I am not attempting to cause you offence, Daniel ?? this is the NCC, this is the process, it's not personal ?? is that yes, you can assume there is consensus but we had no visibility of that, there was nothing in MAT, there was nothing in the mailing list. I realise, Kurtis, we are very tight on time, but, you know, that's why I use that word and that's why I use it in that way.
DANIEL KARRENBERG: The other thing, obviously, is if it turns out there is no consensus here at this meeting, obviously we'll change the plans. That's ?? it's not like, oh, we made this decision, it's irreversible now. But at some point this has been going on for ten years, at some point you have to put a stake in the ground and say this is the time. If it turns out that there are ?? a clear consensus that no, this is not right, then obviously we will change the plans. I don't expect that though.
KURT LINDQVIST: I think that to ?? to sum this up, it's important for the NCC to recognise when you provide services to the community that you do a full audit process and these services are taken serious by the community and I think if Daniel agrees with that that's what he said.
DANIEL KARRENBERG: I agree to that and I take responsibility for the obvious mistake that was made and I pledge to try to not make it again.
AUDIENCE SPEAKER: Jabber comment. I have a comment from Jabber from James Blessing who says that, that the process has been unacceptable.
KURT LINDQVIST: Okay. Any other comments, questions? Thank you Daniel.
(Applause)
KURT LINDQVIST: That means that we are at the very last agenda item, which is open mike. Any last things? Going, going, gone. Please leave the room as quickly as possible. Thank you all for coming. We'll see you at the next one in Amsterdam. Sorry, one more question.
AUDIENCE SPEAKER: This is like open microphone, any other BIS business, right? I have one comment for ?? I think it's almost general to all the presentations we have heard here, and first I'd like to say, I really appreciate and I think that the people in the RIPE NCC are doing ?? making some great tools for our members for the LIRs. But I would like to point out that, I think a little mismatch, or at least of the prioritiation of how the new tools are being developed and what focus and what kind of members' focus of the new tools have. I see that most of the new tools that are being developed are being developed in a website and web applicable and the new, the first versions that comes out are very much focused on the small LIR with not a lot of space that can easily see and develop and work on their space from one website on a page. That fits probably a large majority of the members of the LIRs, but it does not fit with the large majority of the address space that needs to be administered. And I would like to encourage the RIPE NCC and whenever they do their development process and decide what kind of approach do we have to this new tool that we want to make, that you also, in equal measurement look at the amount of space that this tool is supposed to help the member work on and not only look at the number of LIRs that will benefit from the tool and from the first versions of the tool. That was it.
(Applause)
And I do take comments back from the N contribution CC.
DANIEL KARRENBERG: This was about the analyser or tools in general?
AUDIENCE SPEAKER: The analyser is the most important one. I see it in multiple tools...
DANIEL KARRENBERG: I can say for the, for all the tools I think, and you can get into that, Andrew can get into that, we are really looking at quite clearly digitising them and also making the raw data that's in the tools available, so if people want to write their own applications or use widgets and things like that, that's really on our plans and actually for RIPE Stat and the Atlas stuff it's already there I think for the analyser it's planned. So what you want we already plan.
KURT LINDQVIST: Thank you. With that, we'll be done. And now I ask you to leave the room and then we will move to the AGM. Thank you.
LIVE CAPTIONING BY MARY McKEON RPR
DOYLE COURT REPORTERS LTD, DUBLIN, IRELAND.
WWW.DCR.IE