From dns at chasqui.rcp.net.pe Thu Jan 9 23:30:56 1997 From: dns at chasqui.rcp.net.pe (Usuario de estudio de DNS) Date: Thu, 9 Jan 1997 17:30:56 -0500 (EST) Subject: Problems with root-servers Message-ID: Hi On the last days we check my DNS with doc and it report these problems DIGERR (UNKNOWN) : dig @c.root-servers.net. for NS of pe. failed Please could you give me an advise how i can correct this problem ? Thank you From GeertJan.deGroot at ripe.net Fri Jan 10 00:10:19 1997 From: GeertJan.deGroot at ripe.net (Geert Jan de Groot) Date: Fri, 10 Jan 1997 00:10:19 +0100 Subject: Problems with root-servers In-Reply-To: Your message of "Thu, 09 Jan 1997 17:30:56 EST." Message-ID: <9701092310.AA16590@ncc.ripe.net> (note reply-to; I'm not sure this is a Working Group issue) On Thu, 9 Jan 1997 17:30:56 -0500 (EST) Usuario de estudio de DNS wrote: > On the last days we check my DNS with doc and it report these problems > DIGERR (UNKNOWN) : dig @c.root-servers.net. for NS of pe. failed > Please could you give me an advise how i can correct this problem ? Hi, A quick investigation learns that the secondary nameserver for the PE TLD, rain.psg.com, is not carrying your nameserver information. That means that only your nameserver is supplying nameserver information. As I know Randy quite well, I sent a message to him and I'm convinced that he will fix it ASAP. I note there is only one secondary nameserver. The RIPE NCC offers secondary nameserver services for all Top Level Domains, so if you are interested in getting a secondary in Europe, then please contact me privately. Kind regards, Geert Jan de Groot RIPE NCC From Daniel.Karrenberg at ripe.net Mon Jan 13 09:38:51 1997 From: Daniel.Karrenberg at ripe.net (Daniel Karrenberg) Date: Mon, 13 Jan 1997 09:38:51 +0100 Subject: 2nd Root Server in Europe Message-ID: <9701130838.AA16460@ncc.ripe.net> Dear colleagues, I have received an informal query from IANA about RIPE's opinion about the placement of a second root name server in Europe at the LINX operated by VBCnet. This should be discussed at the next meeting. Personally I think that placement at the LINX makes good sense because of current network topology. Daniel ----- Begin Included Message ----- >From jdd at avon-gw.uk1.vbc.net Thu Jan 9 11:03:28 1997 Date: Thu, 9 Jan 1997 19:04:48 +0000 (GMT) From: Jim Dixon To: iana at iana.org, postel at isi.edu Cc: paul at vix.com Subject: Re: UK root nameserver Two months ago I wrote to you regarding the possibility of our setting up a root name server in the UK. To date I have received no reply at all, other than something from a robot promising a reply from a human being within ten days; this of course never came. I have been just now been asked to report back to the board of the UK Internet Service Providers Association regarding this request. Would it be possible for someone at IANA to give us some indication of what your response is likely to be? On Mon, 11 Nov 1996, Jim Dixon wrote: > Several days ago one of our engineers talked to Paul Vixie in California > regarding our setting up a root name server in the United Kingdom. As > I am sure you are aware, there is only one root server in Europe and > for many UK networks that server is netwise further from the UK than > most of the US servers. > > We will have a server built to Paul Vixie's recommendations some time > this week. As he suggests, it will be placed in its own class C on a > private LAN segment. This will be on a 100 Mbps port on a router with > a 100 Mbps connection to the LINX, the London Internet Exchange, which > is the major UK peering point (http://www.linx.net). The LINX is > actually attracting a significant number of networks from the continent: > we now have one Dutch and two German members, with EBONE talking about > joining in December. > > This proposal has the endorsement of the UK Internet Service Providers > Association (http://www.ispa.org.uk), of which I am one of the > directors. ISPA is more or less equivalent to the CIX in the US. The > machine will be operated by VBCnet, which is an owner/ member of the > LINX, and will be, as I said, one hop from the LINX on a 100 Mbps port. > The LINX and our London PoP are located in Telehouse Europe, which is > the UK's premier telecommunications centre. Engineering staff are on duty > 24x7. Telehouse has exceptionally high security, multiply redundant high > speed telecommunications links, and multiply redundant power supplies. > > Should you need any further information or additional material in > support of this proposal, we will of course be happy to supply it. > I can be contacted at either of the numbers below, by email, or > at 011 44 973 737 881 anytime. - -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 ----- End Included Message ----- From keith at linx.org Mon Jan 13 22:58:33 1997 From: keith at linx.org (Keith Mitchell) Date: Mon, 13 Jan 1997 21:58:33 +0000 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <19970113112348.12069.qmail@pool.pipex.net> References: <19970113112348.12069.qmail@pool.pipex.net> Message-ID: <970113215833.ZM21698@stamford.linx.net> Daniel, I've had a number of discussions with Bill Manning to facilitate this, and hosting of this server was again discusssed and endorsed at the LINX meeting today. However, I am rather surprised to hear of the involvement of VBCnet and ISPA - despite Jim regularly attending LINX meetings and being active in LINX mailing lists, he has not mentioned anything about either of these. Normally the ISPA does not concern itself with technical or infrastructure matters. LINX policy, which was endorsed at the two recent IEPG meetings, is that the proposed root server would be in the LINX neutral AS rather than that of any particular member - clearly we welcome VBC's offer of resources, but I am not sure the other LINX members would be quite so supportive of this if they felt one member was being a chance they also might also want. With this qualification LINX remains highly supportive of this proposal and I'll do my best to facilitate it. I will be at the RIPE meeting next Mon and Tue, and hope Jim will be there to clarify any confusion. Keith Mitchell Chairman London InterNet Exchange keith at linx.org PO Box 51 Stamford, PE9 2WF United Kingdom Phone: +44 7000 783797 (fax 783798) Mobile: +44 385 346152 > >To: dns-wg at ripe.net > >Cc: RIPE WG Chairpeople > >Subject: 2nd Root Server in Europe > >From: Daniel Karrenberg > > >I have received an informal query from IANA about RIPE's opinion about > >the placement of a second root name server in Europe at the LINX > >operated by VBCnet. This should be discussed at the next meeting. > >Personally I think that placement at the LINX makes good sense because > >of current network topology. > >On Mon, 11 Nov 1996, Jim Dixon wrote: > > >> We will have a server built to Paul Vixie's recommendations some time > >> this week. As he suggests, it will be placed in its own class C on a > >> private LAN segment. This will be on a 100 Mbps port on a router with > >> a 100 Mbps connection to the LINX, the London Internet Exchange, which > >> is the major UK peering point (http://www.linx.net). The LINX is > >> actually attracting a significant number of networks from the continent: > >> we now have one Dutch and two German members, with EBONE talking about > >> joining in December. > >> > >> This proposal has the endorsement of the UK Internet Service Providers > >> Association (http://www.ispa.org.uk), of which I am one of the > >> directors. ISPA is more or less equivalent to the CIX in the US. The > >> machine will be operated by VBCnet, which is an owner/ member of the > >> LINX, and will be, as I said, one hop from the LINX on a 100 Mbps port. From roll at Stupi.SE Mon Jan 13 23:14:55 1997 From: roll at Stupi.SE (Peter Lothberg) Date: Mon, 13 Jan 97 23:14:55 MET Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of Mon, 13 Jan 1997 21:58:33 +0000 Message-ID: > LINX policy, which was endorsed at the two recent IEPG meetings, is > that the proposed root server would be in the LINX neutral AS This is was was discussed at the IEPG meeting, and the concensus was that "common resources" should be attached to public exchange points trough a dedicated router in a neutral way. Shared resources has to bee keept as neutral as possible, and having for example a root-name-server behind it's own router attached to a well recogized public exchange point seems to be the best avaliable implementation today. Thus all ISP's present at that echange point could peer, regardless of their peering policies to other ISP's and as there is no other service than the nameserver is behind that router, and I'm sure all of them (the ISP's peering) will give transit for the prefix used by the nameserver to other parts of the network. --Peter From jdd at vbc.net Tue Jan 14 01:01:41 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 14 Jan 1997 00:01:41 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <970113215833.ZM21698@stamford.linx.net> Message-ID: On Mon, 13 Jan 1997, Keith Mitchell wrote: > I've had a number of discussions with Bill Manning to facilitate > this, and hosting of this server was again discusssed and endorsed > at the LINX meeting today. I was unable to attend the LINX meeting today, but I was pleased to hear that you had suddenly taken an active interest in the idea of a root name server at the LINX. > However, I am rather surprised to hear of the involvement of VBCnet > and ISPA - despite Jim regularly attending LINX meetings and being > active in LINX mailing lists, he has not mentioned anything about > either of these. Normally the ISPA does not concern itself with > technical or infrastructure matters. Neither the LINX nor ISPA is particularly technical. I am encouraging ISPA to become more involved in technical issues, because I think that it is important to develop UK forums. But the reality is that in both organisations technical issues are given what I believe is too little attention. Last year VBCnet proposed to IANA that we should set up a root name server in the UK because we believed that one was sorely needed and no one else was taking any action on this. ISPA endorsed this action for much the same reasons. There has also been discussion at ISPA about providing other needed technical resources; I think that a little technical competition would inspire us all to move faster ;-) The last time this was discussed at the LINX, the conclusion was that the LINX itself would not take any action until the LINX had hired and trained a systems engineer, if then. You said this yourself, Keith. As I recall, there was a vague suggestion that some action might be initiated in six months. Like, for example, Demon, we have been impatient with the slow pace of things at the LINX. I understand that Demon today announced that they would be bringing up a route server at the LINX. I know that they proposed this many months ago and were met with indifference. So they unilaterally took action. We have been doing the same as regards a root name server. We were not just impatient with the lack of progress. What has been discussed in the past at the LINX, and I believe was discussed again today, was a root server for "." We think that what would be of most benefit to Europe would be a full root name server, including the largest domains (.com and .net), and this is what we discussed with Paul Vixie and proposed to Jon Postel last year. We are not proposing to put up a minimal root name server; frankly, I can't see any purpose in minimizing the name server's scope. We believe that what is needed is a full-blown name server. > LINX policy, which was endorsed at the two recent IEPG meetings, is > that the proposed root server would be in the LINX neutral AS rather > than that of any particular member - clearly we welcome VBC's offer > of resources, but I am not sure the other LINX members would be > quite so supportive of this if they felt one member was being > a chance they also might also want. I don't quite understand what is meant here. We have offered to provide and operate a root name server at Telehouse at our own expense. We originally began looking into this because we wanted a root name server for our own use; then we talked it over and decided to make it available to everyone at no charge as a service to the UK / European Internet community. ISPA endorsed it on this basis. There is a long tradition of this in this country; EUnet GB, for example, operated the .uk name servers on the same basis for years. I think that many or most LINX members would like to see a root name server at the LINX and wouldn't much care who manages it. We will begin formally canvassing other LINX members on this issue tomorrow, but those that we have spoken with so far are supportive. Certainly ISPA members are. > With this qualification LINX remains highly supportive of this > proposal and I'll do my best to facilitate it. We would certainly appreciate this. Given official approval, we can have a server up in days. We were prepared to do so in November. > I will be at the RIPE > meeting next Mon and Tue, and hope Jim will be there to clarify any > confusion. I myself cannot be there; I have critical prior commitments. However VBCnet will be represented and I am certain that we can come to a arrangement that will satisfy all concerned. > Keith Mitchell > > Chairman > London InterNet Exchange keith at linx.org > PO Box 51 > Stamford, PE9 2WF > United Kingdom > Phone: +44 7000 783797 (fax 783798) > Mobile: +44 385 346152 > > > >To: dns-wg at ripe.net > > >Cc: RIPE WG Chairpeople > > >Subject: 2nd Root Server in Europe > > >From: Daniel Karrenberg > > > > >I have received an informal query from IANA about RIPE's opinion about > > >the placement of a second root name server in Europe at the LINX > > >operated by VBCnet. This should be discussed at the next meeting. > > >Personally I think that placement at the LINX makes good sense because > > >of current network topology. > > > >On Mon, 11 Nov 1996, Jim Dixon wrote: > > > > >> We will have a server built to Paul Vixie's recommendations some time > > >> this week. As he suggests, it will be placed in its own class C on a > > >> private LAN segment. This will be on a 100 Mbps port on a router with > > >> a 100 Mbps connection to the LINX, the London Internet Exchange, which > > >> is the major UK peering point (http://www.linx.net). The LINX is > > >> actually attracting a significant number of networks from the continent: > > >> we now have one Dutch and two German members, with EBONE talking about > > >> joining in December. > > >> > > >> This proposal has the endorsement of the UK Internet Service Providers > > >> Association (http://www.ispa.org.uk), of which I am one of the > > >> directors. ISPA is more or less equivalent to the CIX in the US. The > > >> machine will be operated by VBCnet, which is an owner/ member of the > > >> LINX, and will be, as I said, one hop from the LINX on a 100 Mbps port. > -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From jdd at vbc.net Tue Jan 14 09:24:14 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 14 Jan 1997 08:24:14 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Message-ID: On Mon, 13 Jan 1997, Peter Lothberg wrote: > This is was was discussed at the IEPG meeting, and the concensus was > that "common resources" should be attached to public exchange points > trough a dedicated router in a neutral way. > > Shared resources has to bee keept as neutral as possible, and having > for example a root-name-server behind it's own router attached to a well > recogized public exchange point seems to be the best avaliable > implementation today. This is in fact exactly the physical arrangement that we have proposed: a root name server on its own class C one hop off the LINX. And the UK Internet Service Providers Asssociation (http://www.ispa.org) is as neutral an organisation as the LINX -- in fact arguably more so. ISPA draws its membership from the entire UK Internet community, whereas LINX's membership rules are quite restrictive. This should not be overemphasized; there is a considerable overlap in membership between the two organisations. Like many other networks, VBCnet belongs to both. > Thus all ISP's present at that echange point could peer, regardless of > their peering policies to other ISP's and as there is no other service > than the nameserver is behind that router, and I'm sure all of them > (the ISP's peering) will give transit for the prefix used by the > nameserver to other parts of the network. We agree with this entirely. To repeat what may not be clear to those added to the lengthening CC: list, some time ago, after securing the endorsement of the UK ISPA, VBCnet approached IANA on behalf of ISPA with an offer to provide and manage a root name server at the LINX. We sent an engineer to California to discuss this with Paul Vixie, who operates one of the root name servers and is responsible for the software. We intend to follow Vixie's recommendations on hardware, software, and the way in which the root name server is attached to the LINX, which is in fact as you have described above. -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From Daniel.Karrenberg at ripe.net Tue Jan 14 09:46:38 1997 From: Daniel.Karrenberg at ripe.net (Daniel Karrenberg) Date: Tue, 14 Jan 1997 09:46:38 +0100 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of Mon, 13 Jan 1997 21:58:33 GMT. <970113215833.ZM21698@stamford.linx.net> References: <970113215833.ZM21698@stamford.linx.net> Message-ID: <9701140846.AA16493@ncc.ripe.net> Keith, thank you for your message. I personally agree that topologically speaking the LINX is an excellent place for a 2nd root nameserver. I also agree that from a routing standpoint such a box should not be in any particular ISP's AS and it should not be operated by a particular ISP either unless there is very wide and stable consensus that this is the right thing to do. At the RIPE meeting I would like to establish consensus about the physical location (LINX, telehouse), the routing location (LINX AS or dedicated AS) and the operations. About the latter I personally think that operation by the RIPE NCC in cooperation with the LINX is a good soloution. While I may be somewhat biased in favour of the NCC ;-), I think this is good because it would allow all ISPs in the region to buy into the soloution rather than only the LINX members. After all the NCC's role in life is to be the neutral place where all ISPs in the region can do things they need to do together. Do I make sense? Daniel From keith at linx.org Tue Jan 14 12:31:34 1997 From: keith at linx.org (Keith Mitchell) Date: Tue, 14 Jan 1997 11:31:34 +0000 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: References: Message-ID: <970114113135.ZM22117@stamford.linx.net> Jim, In , wrote: > Neither the LINX nor ISPA is particularly technical. If the LINX is not technical, how come we are running a sucessful exchange point ? > The last time this was discussed at the LINX, the conclusion was > that the LINX itself would not take any action until the LINX had > hired and trained a systems engineer, if then. You said this > yourself, Keith. As I recall, there was a vague suggestion that > some action might be initiated in six months. There have been some developments on this subject to further discussions with ISI - the situation has changed. > Like, for example, Demon, we have been impatient with the slow pace > of things at the LINX. I understand that Demon today announced that > they would be bringing up a route server at the LINX. I know that > they proposed this many months ago and were met with indifference. > So they unilaterally took action. We have been doing the same as > regards a root name server. This is a mis-representation of Demon's position at yesterday's meeting that you did not attend - their action is co-operative rather than unilateral. I am sure Demon will correct whichever of us they feel is mis-representing them. > What has been > discussed in the past at the LINX, and I believe was discussed again > today, was a root server for "." We think that what would be of most > benefit to Europe would be a full root name server, including the > largest domains (.com and .net), and this is what we discussed with > Paul Vixie and proposed to Jon Postel last year. > > We are not proposing to put up a minimal root name server; frankly, > I can't see any purpose in minimizing the name server's scope. > We believe that what is needed is a full-blown name server. I disagree, but let's debate this in the relevant forums first, and not go off and do our thing. In , wrote: > On Mon, 13 Jan 1997, Peter Lothberg wrote: > > Shared resources has to bee keept as neutral as possible, and having > > for example a root-name-server behind it's own router attached to a well > > recogized public exchange point seems to be the best avaliable > > implementation today. > > This is in fact exactly the physical arrangement that we have proposed: > a root name server on its own class C one hop off the LINX. Not from a routing policy point of view - will you put the root server in a seperate AS, or will connectivity be constrained by your peering/routing policy ? > To repeat what may not be clear to those > added to the lengthening CC: list, some time ago, after securing the > endorsement of the UK ISPA, VBCnet approached IANA on behalf of ISPA > with an offer to provide and manage a root name server at the LINX. Neither VBCnet nor ISPA has the authority to offer "to provide and manage a root name server at the LINX", without appropriate approval from a majority of LINX members. They do of course have authority to offer to provide and manage a root nameserver on their own, or another of their members' own network. > We sent an engineer to California to discuss this with Paul Vixie, > who operates one of the root name servers and is responsible for the > software. We intend to follow Vixie's recommendations on hardware, > software, and the way in which the root name server is attached to > the LINX, which is in fact as you have described above. Could someone maybe reconcile this with the offer of ready hardware I've had from Bill Manning ? I'm confused. I think we need a little more discussion between those who have plans and policy ideas here, and a rather less attempts at public point-scoring to the widest possible distribution. Keith Mitchell Chairman London InterNet Exchange keith at linx.org PO Box 51 Stamford, PE9 2WF United Kingdom Phone: +44 7000 783797 (fax 783798) Mobile: +44 385 346152 From jdd at vbc.net Tue Jan 14 12:46:38 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 14 Jan 1997 11:46:38 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <970114113135.ZM22117@stamford.linx.net> Message-ID: On Tue, 14 Jan 1997, Keith Mitchell wrote: > > Neither the LINX nor ISPA is particularly technical. > > If the LINX is not technical, how come we are running a sucessful > exchange point ? Is this worth going into? Yes we cooperate in operating a successful exchange point at the LINX. But the LINX is not a centre for new ideas; I personally would prefer that it was. > > The last time this was discussed at the LINX, the conclusion was > > that the LINX itself would not take any action until the LINX had > > hired and trained a systems engineer, if then. You said this > > yourself, Keith. As I recall, there was a vague suggestion that > > some action might be initiated in six months. > > There have been some developments on this subject to further > discussions with ISI - the situation has changed. So I understand. However, to make things clear, your discussion with Bill Manning followed our approach to IANA by quite some time. > This is a mis-representation of Demon's position at yesterday's > meeting that you did not attend - their action is co-operative > rather than unilateral. I am sure Demon will correct whichever of us > they feel is mis-representing them. I stand ready to be chastised ;-) > > What has been > > discussed in the past at the LINX, and I believe was discussed again > > today, was a root server for "." We think that what would be of most > > benefit to Europe would be a full root name server, including the > > largest domains (.com and .net), and this is what we discussed with > > Paul Vixie and proposed to Jon Postel last year. > > > > We are not proposing to put up a minimal root name server; frankly, > > I can't see any purpose in minimizing the name server's scope. > > We believe that what is needed is a full-blown name server. > > I disagree, but let's debate this in the relevant forums first, and > not go off and do our thing. We would appear to be in the relevant forums. -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From tonyb at pipex.net Tue Jan 14 13:37:55 1997 From: tonyb at pipex.net (Tony Barber) Date: Tue, 14 Jan 1997 12:37:55 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: from "Jim Dixon" at Jan 14, 97 08:24:14 am Message-ID: <19970114123755.5222.qmail@pool.pipex.net> Jim Dixon wrote: > >> Thus all ISP's present at that echange point could peer, regardless of >> their peering policies to other ISP's and as there is no other service >> than the nameserver is behind that router, and I'm sure all of them >> (the ISP's peering) will give transit for the prefix used by the >> nameserver to other parts of the network. > >We agree with this entirely. To repeat what may not be clear to those >added to the lengthening CC: list, some time ago, after securing the >endorsement of the UK ISPA, VBCnet approached IANA on behalf of ISPA I'm sorry but there isn't and never has been any requirement to enforce peering at any exchange point other then for instance CIX. VBC by housing a valuable resource are saying in effect 'yes you can use this root server if you peer with us'. This is not really acceptable. Regards --Tony From jdd at vbc.net Tue Jan 14 13:52:30 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 14 Jan 1997 12:52:30 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <19970114123755.5222.qmail@pool.pipex.net> Message-ID: On Tue, 14 Jan 1997, Tony Barber wrote: > >We agree with this entirely. To repeat what may not be clear to those > >added to the lengthening CC: list, some time ago, after securing the > >endorsement of the UK ISPA, VBCnet approached IANA on behalf of ISPA > > I'm sorry but there isn't and never has been any requirement to enforce > peering at any exchange point other then for instance CIX. > VBC by housing a valuable resource are saying in effect 'yes you can > use this root server if you peer with us'. > This is not really acceptable. This is a complete misreading of what we have said. On the one hand we peer with everyone at the LINX, with one exception, MITE, whose routes we already see through GSL. We have told MITE that if and when they have routes that are not visible through them, we will peer with them. So we aren't forcing anyone to do anything. On the other hand, following recommendations to us, our intention is to put a root name server in a class C by itself. There has been a lot of discussion in this area, and this is what I understand to be the recommended policy. We don't care what AS this class C is in; on the whole we would probably prefer that it not be in one of our ASs. -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From jdd at vbc.net Tue Jan 14 14:13:22 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 14 Jan 1997 13:13:22 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <970114113135.ZM22117@stamford.linx.net> Message-ID: On Tue, 14 Jan 1997, Keith Mitchell wrote: > > On Mon, 13 Jan 1997, Peter Lothberg wrote: > > > > Shared resources has to bee keept as neutral as possible, and having > > > for example a root-name-server behind it's own router attached to a well > > > recogized public exchange point seems to be the best avaliable > > > implementation today. > > > > This is in fact exactly the physical arrangement that we have proposed: > > a root name server on its own class C one hop off the LINX. > > Not from a routing policy point of view - will you put the root > server in a seperate AS, or will connectivity be constrained by > your peering/routing policy ? As I think that I have already said, we don't much care which AS it is in, and, no, it would not be constrained by our peering/routing policy. > > To repeat what may not be clear to those > > added to the lengthening CC: list, some time ago, after securing the > > endorsement of the UK ISPA, VBCnet approached IANA on behalf of ISPA > > with an offer to provide and manage a root name server at the LINX. > > Neither VBCnet nor ISPA has the authority to offer "to provide and > manage a root name server at the LINX", without appropriate approval > from a majority of LINX members. They do of course have authority to > offer to provide and manage a root nameserver on their own, or > another of their members' own network. I was quickly restating our offer and did so sloppily. If you look back at what I was restating, it used the phrase "one hop away from the LINX". I believe that everything else that I have written has made the correct distinctions. VBCnet is a member of the LINX. We and other members use its facilities only by mutual agreement. All of us understand that perfectly clearly. > > We sent an engineer to California to discuss this with Paul Vixie, > > who operates one of the root name servers and is responsible for the > > software. We intend to follow Vixie's recommendations on hardware, > > software, and the way in which the root name server is attached to > > the LINX, which is in fact as you have described above. > > Could someone maybe reconcile this with the offer of ready hardware > I've had from Bill Manning ? I'm confused. Sorry, but I know nothing about this. > I think we need a little > more discussion between those who have plans and policy ideas here, > and a rather less attempts at public point-scoring to the widest > possible distribution. Let us make this also perfectly clear: I have added not one name to the CC list. I did correct "chief at ispa.org.uk" to "ispa-com at ispa.org.uk". It is others who have grown and grown the CC list and are, if you are correct, interested in "public point-scoring to the widest possible distibution". -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From bmanning at ISI.EDU Tue Jan 14 16:45:25 1997 From: bmanning at ISI.EDU (bmanning at ISI.EDU) Date: Tue, 14 Jan 1997 07:45:25 -0800 (PST) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <970114113135.ZM22117@stamford.linx.net> from "Keith Mitchell" at Jan 14, 97 11:31:34 am Message-ID: <199701141545.AA10548@zed.isi.edu> > > We sent an engineer to California to discuss this with Paul Vixie, > > who operates one of the root name servers and is responsible for the > > software. We intend to follow Vixie's recommendations on hardware, > > software, and the way in which the root name server is attached to > > the LINX, which is in fact as you have described above. > > Could someone maybe reconcile this with the offer of ready hardware > I've had from Bill Manning ? I'm confused. I think we need a little > more discussion between those who have plans and policy ideas here, > and a rather less attempts at public point-scoring to the widest > possible distribution. > I would expect that Paul Vixies recommendations fall closely in line with RFC 2010 and I would expect that any new root nameservers would fall under this document as far as operational activities are concerned. I will note that neither Paul or myself has the final say in this matter. One school of thought is for the IANA to provide the hardware the other is to have the organization provide it. The actual attachment would seem to be best served off a public exchange with the machine on its own prefix and behind its own AS. --bill From jdd at vbc.net Tue Jan 14 19:55:43 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 14 Jan 1997 18:55:43 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <199701141545.AA10548@zed.isi.edu> Message-ID: On Tue, 14 Jan 1997 bmanning at ISI.EDU wrote: > > > We sent an engineer to California to discuss this with Paul Vixie, > > > who operates one of the root name servers and is responsible for the > > > software. We intend to follow Vixie's recommendations on hardware, > > > software, and the way in which the root name server is attached to > > > the LINX, which is in fact as you have described above. > > ... > I would expect that Paul Vixies recommendations fall closely in > line with RFC 2010 and I would expect that any new root nameservers > would fall under this document as far as operational activities > are concerned. Just for the record, we had someone talk to Paul Vixie because RFC 2010 lacks sufficient detail. Yes, his recommendations followed the RFC, but they were much clearer and more specific. > The actual attachment would seem to be best served off a public > exchange with the machine on its own prefix and behind its own AS. Understood and agreed. There is no contradiction between this and our original proposal. -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From amb at xara.net Tue Jan 14 19:05:41 1997 From: amb at xara.net (Alex.Bligh) Date: Tue, 14 Jan 1997 18:05:41 +0000 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of "Tue, 14 Jan 1997 12:52:30 GMT." Message-ID: <199701141805.SAA27925@diamond.xara.net> > On the other hand, following recommendations to us, our intention is > to put a root name server in a class C by itself. There has been a > lot of discussion in this area, and this is what I understand to be > the recommended policy. We don't care what AS this class C is in; > on the whole we would probably prefer that it not be in one of our > ASs. At the risk of AOL-ism, I too prefer the consensus agreement reached by all the LINX members who were at the AGM. Just because all LINX members currently peer with VBC, this does not mean that new members will be allowed to, want to, or that current members will continue to be allowed to / continue to want to. As I understand it, the latest IETF & IEPG proposals suggest that the servers should be at exchange points transitted by several people and run in a neutral manner (apologies if this a misrepresentation), not transitted by a single member, albeit from a point close to an IXP. As far as I know the LINX proposal has always been on the table modulo the issue of whether .com etc. where secondaried there as well as '.' and the transit bandwidth implications thereof. Alex Bligh Xara Networks From gih at telstra.net Wed Jan 15 01:28:45 1997 From: gih at telstra.net (Geoff Huston) Date: Wed, 15 Jan 1997 10:28:45 +1000 Subject: 2nd Root Server in Europe (fwd) Message-ID: This was discussed at the IEPG meeting in July and again in December. The issue at the time was to provide advice to IANA as to the potential siting of additional root name servers - the advice so provided was along the lines of siting one server in the UK with a potential system host of LINX and siting a second in Japan, with a potential host of WIDE. The discussion was initiated by Bill Manning and the notes of the Montreal meeting are at http://www.iepg.org Thanks Geoff From schneider at switch.ch Wed Jan 15 16:19:09 1997 From: schneider at switch.ch (Marcel Schneider) Date: Wed, 15 Jan 1997 16:19:09 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Message from Daniel Karrenberg of "Mon, 13 Jan 1997 09:38:51 +0100." <9701130838.AA16460@ncc.ripe.net> References: <9701130838.AA16460@ncc.ripe.net> Message-ID: <9701151519.AA29426@ncc.ripe.net> On Monday, 13 Jan 1997, Daniel Karrenberg writes: > Dear colleagues, > I have received an informal query from IANA about RIPE's opinion about > the placement of a second root name server in Europe at the LINX > operated by VBCnet. This should be discussed at the next meeting. > Personally I think that placement at the LINX makes good sense because > of current network topology. Is that true ? We currently have ; formerly NIC.NORDU.NET ; . 3600000 NS I.ROOT-SERVERS.NET. I.ROOT-SERVERS.NET. 3600000 A 192.36.148.17 in the north of Europe and according to above proposal a second root name server in the north of Europe, just one hop from one to the other. How about middle and south Europe ? Wouldn't it make sense to have at least one additional NS in these densely populated regions to provide better access times there ? This is just a question, nothing more. > Daniel Marcel Schneider at SWITCH > ----- Begin Included Message ----- >>From jdd at avon-gw.uk1.vbc.net Thu Jan 9 11:03:28 1997 > Date: Thu, 9 Jan 1997 19:04:48 +0000 (GMT) > From: Jim Dixon > To: iana at iana.org, postel at isi.edu > Cc: paul at vix.com > Subject: Re: UK root nameserver > Two months ago I wrote to you regarding the possibility of our setting > up a root name server in the UK. To date I have received no reply at > all, other than something from a robot promising a reply from a human > being within ten days; this of course never came. > I have been just now been asked to report back to the board of the UK > Internet Service Providers Association regarding this request. Would > it be possible for someone at IANA to give us some indication of what > your response is likely to be? > On Mon, 11 Nov 1996, Jim Dixon wrote: >> Several days ago one of our engineers talked to Paul Vixie in California >> regarding our setting up a root name server in the United Kingdom. As >> I am sure you are aware, there is only one root server in Europe and >> for many UK networks that server is netwise further from the UK than >> most of the US servers. >> >> We will have a server built to Paul Vixie's recommendations some time >> this week. As he suggests, it will be placed in its own class C on a >> private LAN segment. This will be on a 100 Mbps port on a router with >> a 100 Mbps connection to the LINX, the London Internet Exchange, which >> is the major UK peering point (http://www.linx.net). The LINX is >> actually attracting a significant number of networks from the continent: >> we now have one Dutch and two German members, with EBONE talking about >> joining in December. >> >> This proposal has the endorsement of the UK Internet Service Providers >> Association (http://www.ispa.org.uk), of which I am one of the >> directors. ISPA is more or less equivalent to the CIX in the US. The >> machine will be operated by VBCnet, which is an owner/ member of the >> LINX, and will be, as I said, one hop from the LINX on a 100 Mbps port. >> The LINX and our London PoP are located in Telehouse Europe, which is >> the UK's premier telecommunications centre. Engineering staff are on duty >> 24x7. Telehouse has exceptionally high security, multiply redundant high >> speed telecommunications links, and multiply redundant power supplies. >> >> Should you need any further information or additional material in >> support of this proposal, we will of course be happy to supply it. >> I can be contacted at either of the numbers below, by email, or >> at 011 44 973 737 881 anytime. > - -- > Jim Dixon VBCnet GB Ltd http://www.vbc.net > tel +44 117 929 1316 fax +44 117 927 2015 > ----- End Included Message ----- From jdd at vbc.net Wed Jan 15 16:24:05 1997 From: jdd at vbc.net (Jim Dixon) Date: Wed, 15 Jan 1997 15:24:05 +0000 (GMT) Subject: subscribe Message-ID: subscribe From rv at NIC.DTAG.DE Thu Jan 16 02:10:04 1997 From: rv at NIC.DTAG.DE (Ruediger Volk) Date: Thu, 16 Jan 1997 02:10:04 +0100 Subject: proposed agenda - DNS WG session at RIPE26 Message-ID: <14718.853377004@kronos.NIC.DTAG.DE> Dear colleagues, please find below a proposed agenda for a DNS WG session at the upcoming RIPE26. 1. Opening & administrativa 2. Agenda bashing 3. Report from San Jose IETF (IRE(?) BOF) 4. Report from IAHC proposal debate (mailing list) 5. Collect & prepare RIPE feedback to IAHC 6. proposed root server from VBCnet 7. technical DNS issues (a) Status new DNS standards work (dnssec, dnsind) (b) new software status (c) any recent technical problems to discuss? 8. reports/news from various top level domains 9. AOB For the reporting items (3, 4, 7*, 8) I would like to ask for volunteers who participated or observed the events or processes to identify to me the kind offer to present a short summary report or position. I have included items 7. and 8. but consider those optional; if neither reports are offered nor priority requests for discussing specific issues are received I think these items should be dropped - as time will be short for the remaining items anyway. Ruediger From roll at Stupi.SE Sun Jan 19 16:25:47 1997 From: roll at Stupi.SE (Peter Lothberg) Date: Sun, 19 Jan 97 16:25:47 MET Subject: 2nd Root Server in Europe In-Reply-To: Your message of Wed, 15 Jan 1997 16:19:09 +0100 Message-ID: > > Dear colleagues, > > > I have received an informal query from IANA about RIPE's opinion about > > the placement of a second root name server in Europe at the LINX > > operated by VBCnet. This should be discussed at the next meeting. > > Personally I think that placement at the LINX makes good sense because > > of current network topology. > > Is that true ? We currently have > > ; formerly NIC.NORDU.NET > ; > . 3600000 NS I.ROOT-SERVERS.NET. > I.ROOT-SERVERS.NET. 3600000 A 192.36.148.17 > > in the north of Europe and according to above proposal a second > root name server in the north of Europe, just one hop from one > to the other. How about middle and south Europe ? Wouldn't it > make sense to have at least one additional NS in these densely > populated regions to provide better access times there ? > > This is just a question, nothing more. > > > > Daniel > > > Marcel Schneider at SWITCH The geographic map does not apply, we have to look at the network topology map. (And please forget the political map to.) If you look at network topology, you will find that the center of Europe is somwhere between NY and Washington DC, followed by Stockholm and London. --Peter From roll at Stupi.SE Sun Jan 19 16:00:46 1997 From: roll at Stupi.SE (Peter Lothberg) Date: Sun, 19 Jan 97 16:00:46 MET Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of Tue, 14 Jan 1997 11:31:34 +0000 Message-ID: > > What has been > > discussed in the past at the LINX, and I believe was discussed again > > today, was a root server for "." We think that what would be of most > > benefit to Europe would be a full root name server, including the > > largest domains (.com and .net), and this is what we discussed with > > Paul Vixie and proposed to Jon Postel last year. The .COM and .NET domains are very large, they have many querries, but the . nameservers serves the universe. Suggestion; Separate the . (root) from the next level (.country, .net, .com) servers.. So, instead of putting ONE host on the network behind the router, put TWO, and the second one could handle the second level domains you wish. Maybe there should be a THIRD host to serve the country toplevel domain (.uk in this case) and maybe toplevel for other countries the region. (.nl, .es, etc..) Hosts are cheap, running them with a quality service is more expensive, but the number of users affected by the performance and availiabilty and performace are VERY high... > > > Shared resources has to bee keept as neutral as possible, and having > > > for example a root-name-server behind it's own router attached to a well > > > recogized public exchange point seems to be the best avaliable > > > implementation today. > > > > This is in fact exactly the physical arrangement that we have proposed: > > a root name server on its own class C one hop off the LINX. > > Not from a routing policy point of view - will you put the root > server in a seperate AS, or will connectivity be constrained by > your peering/routing policy ? The nameservers has to be in their own AS, behind their own router(s) and account for traffic to be in the 400kbps in / 700kbps out range and estimate growth to 10% month. So full-duplex FDDI or full-duplex fast Ethernet is a requrement. The peering policy of this as should be something like; Peer with anyone who request to peer. Maybe it even should accept circiuts from those who just want to have good connectivity to the nameservers, but don't care about the particular exchanage point. ------------------------------------------------------------------------ I would suggest that once everyone agreed on a workable model that we also implement that for other root namseservers. --Peter From roll at Stupi.SE Sun Jan 19 16:09:00 1997 From: roll at Stupi.SE (Peter Lothberg) Date: Sun, 19 Jan 97 16:09:00 MET Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of Tue, 14 Jan 1997 13:13:22 +0000 (GMT) Message-ID: > Let us make this also perfectly clear: I have added not one name to > the CC list. > > I did correct "chief at ispa.org.uk" to "ispa-com at ispa.org.uk". > > It is others who have grown and grown the CC list and are, if you are > correct, interested in "public point-scoring to the widest possible > distibution". > Jim Dixon VBCnet GB Ltd http://www.vbc.net > tel +44 117 929 1316 fax +44 117 927 2015 I added most of the CC lines, as I'm very concerned about how we handle limited and shared resources in the global Internet. The purpose of a second european rootnameserver is to improve the global Internet for everyone, not only local people covered by the London metro. --Peter From knm at dial.pipex.com Mon Jan 20 14:08:23 1997 From: knm at dial.pipex.com (Keith N Mitchell) Date: Mon, 20 Jan 1997 13:08:23 +0000 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Peter Lothberg "Re: 2nd Root Server in Europe (fwd)" (Jan 19, 4:00pm) References: Message-ID: <970120130955.ZM9463@boiler.linx.org> Taking a step back from the detailed discussion, it seems we don't have any major objections from the UK community to the proposals for deploying this I circulated last week, so I feel happy to take them to the RIPE DNS WG to take them a step further. We have offers of resources from VBCnet and NOMINET in the form of server hardware and operational support, and various LINX members as regards transit. We have support in principle from LINX, ISPA & NOMINET. Here's what I think the outstanding issues are: - fitting it all in to the European and Global picture: - for further discussion - routing policy for the NS ? various points here: - do we put the server in the LINX AS, or create a new one ? For this point it is worth remembering the reason we have the LINX AS in the first place, namely: 1. To make the LINX LAN prefix visible to the rest of the world 2. To give the LINX secretariat Internet connectivity 3. To have something everyone can peer with to gather useful stats in the collector. I don't think there are any sacred cows for any of these here - if we adjust the LINX policy to meet requirements of the root NS rather than the other way around, I don't think we break either of 1 or 2. I don't think it breaks 3. either, but this requirement will go away when we have a route server in any case. i.e. I am very appreciative of the existing LINX members who give transit to the LINX, but, it would not break my heart to have the same connectivity/policy for the LINX as a UK-based root NS, and would understand if the existing providers did not want to fully transit a root servers' traffic - we have had enough offers to transit the root NS I do not see this as a problem. - how many members should provide transit to the root NS AS ? we have appear to have 2 schools of thought here, namely that either everyone should transit the thing to spread the load as evenly as possible, or that we should only have a small number to keep the routing deterministic (there's also a 3rd more wacky suggestion to do some special local hack, but I tend to agree with Paul Vixie's comments on that). My view is that we have to be careful about too many members providing transit, as we can easily finish up with lots of views of the way to the LINX AS with equal path lengths, but non-obvious tie-breaks, leading to hard to predict connectivity. Also increasing the number of path/prefix products in the global routing tables is in any case ecologically unsound. Perhaps the solution is for a small number of well-connected LINX members to provide universal transit, but for an additional group to provide additional transit to specific places (like the Nacamar offer). (Further discussion of server routing policy is probably best confined to technical forums, i.e. maybe nom-dir and ispa-com want to be dropped off the distribution.) - server for just "." or TLDs ? This kind of ties in with routing policy, in that those providing transit have potential concerns about traffic levels generated by a TLD server . One solution here might be PeterL's suggestion of seperate servers for TLD and ".", with different members providing different transit to each. The TLD server does not need to have its own independent address space in the same way the "." one does. I think we can still get away with having them in the same AS, though there is some risk of asymmetry here. All we need is an additional offer of server hardware..... - router hardware We do need to think about upgrading the collector to a 64Mb- capable box, most likely a 7200. If we think that a 100Mb connection for the server is needed, then the is a spare slot in the collector for an FDDI or 100baseT card. I'm sure there was another point to make, but being in the RIPE terminal room is distracting me a bit. More discussion of this here tomorrow. Keith From bmanning at ISI.EDU Tue Jan 21 06:22:56 1997 From: bmanning at ISI.EDU (Bill Manning) Date: Mon, 20 Jan 1997 21:22:56 -0800 (PST) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <970120130955.ZM9463@boiler.linx.org> from "Keith N Mitchell" at Jan 20, 97 01:08:23 pm Message-ID: <199701210522.AA17994@zephyr.isi.edu> > - server for just "." or TLDs ? > I understand that the offer on the table is just for "." NSI will have to make its own arraingments for .COM/.NET/.ORG secondaries. My personal opinion is to host such servers on their own prefixes. The debates wrt home AS are effectivly moot from my perspective, it can be done a number of ways and not really matter. -- --bill From schneider at switch.ch Tue Jan 21 14:21:49 1997 From: schneider at switch.ch (Marcel Schneider) Date: Tue, 21 Jan 1997 14:21:49 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Message from Peter Lothberg of "Sun, 19 Jan 1997 16:25:47 +0700." References: Message-ID: <9701211322.AA18627@ncc.ripe.net> On Sunday, 19 Jan 1997, Peter Lothberg writes: >> > Dear colleagues, >> >> > I have received an informal query from IANA about RIPE's opinion about >> > the placement of a second root name server in Europe at the LINX >> > operated by VBCnet. This should be discussed at the next meeting. >> > Personally I think that placement at the LINX makes good sense because >> > of current network topology. >> >> Is that true ? We currently have >> >> ; formerly NIC.NORDU.NET >> ; >> . 3600000 NS I.ROOT-SERVERS.NET. >> I.ROOT-SERVERS.NET. 3600000 A 192.36.148.17 >> >> in the north of Europe and according to above proposal a second >> root name server in the north of Europe, just one hop from one >> to the other. How about middle and south Europe ? Wouldn't it >> make sense to have at least one additional NS in these densely >> populated regions to provide better access times there ? >> >> This is just a question, nothing more. >> >> >> > Daniel >> >> >> Marcel Schneider at SWITCH > The geographic map does not apply, we have to look at the network > topology map. (And please forget the political map to.) > If you look at network topology, you will find that the center of > Europe is somwhere between NY and Washington DC, followed by Stockholm > and London. Ah, the center of Europe is between NY and Washington DC and that is reason enough to have _yet another server_ there. Great. Can imagine that you would like even more root name servers there in oder to be able to offer the best connectivity on earth to the Elks and forget about the rest of the world :-). IMHO: Your answer is insufficient if not entirely mistaken. The inhabitants of central and southern Europe have a right to the same kind of connectivity as the northern region is now trying to get hold of or already has done so. Make no sense to improve an already good situation. > --Peter Marcel From wolf at pasteur.fr Tue Jan 21 15:59:54 1997 From: wolf at pasteur.fr (Christophe Wolfhugel) Date: Tue, 21 Jan 1997 15:59:54 +0100 (MET) Subject: 2nd Root Server in Europe In-Reply-To: <9701211322.AA18627@ncc.ripe.net> from Marcel Schneider at "Jan 21, 97 02:21:49 pm" Message-ID: <199701211459.PAA17483@yseult.sis.pasteur.fr> Marcel Schneider: > IMHO: Your answer is insufficient if not entirely mistaken. The inhabitants > of central and southern Europe have a right to the same kind of connectivity > as the northern region is now trying to get hold of or already has done > so. Make no sense to improve an already good situation. What about having the Europeans to work towards moving the center of Europe from the US East coast to some European point ? As long as European major connectivity providers are ignoring each other and having their best interconnection points being MAE-East or the NY Nap all this seems useless to me... (that was an end user's point of view -- intra Europe networking is not usable today for people who work, at least those not in scandinavian areas). So for us, London, Stockholm, NY, Washington, Geneva, Milano, etc... that root server will mainly be unrechable :(. The location of that server is not a technical issue, it's a political one. -- Christophe Wolfhugel -+- SIS, Institut Pasteur, Paris Boulot : wolf at pasteur.fr, $HOME : wolf at schnok.fr.net From schneider at switch.ch Tue Jan 21 16:56:58 1997 From: schneider at switch.ch (Marcel Schneider) Date: Tue, 21 Jan 1997 16:56:58 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Message from Christophe Wolfhugel of "Tue, 21 Jan 1997 15:59:54 +0100." <199701211459.PAA17483@yseult.sis.pasteur.fr> References: <199701211459.PAA17483@yseult.sis.pasteur.fr> Message-ID: <9701211557.AA04082@ncc.ripe.net> On Tuesday, 21 Jan 1997, Christophe Wolfhugel writes: > Marcel Schneider: >> IMHO: Your answer is insufficient if not entirely mistaken. The inhabitants >> of central and southern Europe have a right to the same kind of connectivity >> as the northern region is now trying to get hold of or already has done >> so. Make no sense to improve an already good situation. > What about having the Europeans to work towards moving the center of > Europe from the US East coast to some European point ? > As long as European major connectivity providers are ignoring each other > and having their best interconnection points being MAE-East or the NY Nap > all this seems useless to me... > (that was an end user's point of view -- intra Europe networking is not > usable today for people who work, at least those not in scandinavian areas). > So for us, London, Stockholm, NY, Washington, Geneva, Milano, etc... that > root server will mainly be unrechable :(. > The location of that server is not a technical issue, it's a political one. Agreements, policies. Also my (personal) opinion. But as long as Peter Lothberg et al are attempting to tell us 'route your inter-European IP packets over the US' nothing will improve. The owners of 150+ MB links to the US can easily sell this opinion, but it is absurd. > -- > Christophe Wolfhugel -+- SIS, Institut Pasteur, Paris > Boulot : wolf at pasteur.fr, $HOME : wolf at schnok.fr.net Marcel From kissg at sztaki.hu Tue Jan 21 18:09:50 1997 From: kissg at sztaki.hu (Gabor Kiss) Date: Tue, 21 Jan 1997 18:09:50 +0100 Subject: 2nd Root Server in Europe Message-ID: <199701211709.Co12186@sztaki.hu> Just a note: If anyone tell an address form the network what is the planned location of the new root server, all of us could test the good connectivity. E.g. with traceroute. Regards Gabor From jdd at vbc.net Tue Jan 21 20:01:40 1997 From: jdd at vbc.net (Jim Dixon) Date: Tue, 21 Jan 1997 19:01:40 +0000 (GMT) Subject: 2nd Root Server in Europe In-Reply-To: <199701211709.Co12186@sztaki.hu> Message-ID: On Tue, 21 Jan 1997, Gabor Kiss wrote: > If anyone tell an address form the network > what is the planned location of the new root server, > all of us could test the good connectivity. > E.g. with traceroute. Try collector.linx.net = 194.68.130.254 However, quite a few networks have offered to supply transit to the new root name server, so if anything connectivity should be better than what it is now. -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From pete at sms.fi Tue Jan 21 19:45:27 1997 From: pete at sms.fi (Petri Helenius) Date: Tue, 21 Jan 1997 20:45:27 +0200 (EET) Subject: 2nd Root Server in Europe Message-ID: <199701211845.UAA07430@silver.sms.fi> At 14:21 21.1.1997 +0100, Marcel Schneider wrote: >IMHO: Your answer is insufficient if not entirely mistaken. The inhabitants >of central and southern Europe have a right to the same kind of connectivity >as the northern region is now trying to get hold of or already has done >so. Make no sense to improve an already good situation. > HAVE THE RIGHT? Excuse me? Just go to your telco and order the pipes and stop whining. If they don't, tell everybody that they aren't up to their job. Tell your legistrators that they de-regulate so somebody that _IS_ up to the job can come in and do it properly. Pete From wolf at pasteur.fr Wed Jan 22 08:11:05 1997 From: wolf at pasteur.fr (Christophe Wolfhugel) Date: Wed, 22 Jan 1997 08:11:05 +0100 (MET) Subject: 2nd Root Server in Europe In-Reply-To: from Jim Dixon at "Jan 21, 97 07:01:40 pm" Message-ID: <199701220711.IAA27512@yseult.sis.pasteur.fr> Jim Dixon: > Try collector.linx.net = 194.68.130.254 Bingo ! Still the end user's point of view : traceroute to collector.linx.net (194.68.130.253), 30 hops max, 40 byte packets 1 sisint.net.pasteur.fr (157.99.60.1) 1 ms 1 ms 1 ms 2 cerbere.pasteur.fr (157.99.64.1) 2 ms 1 ms 2 ms 3 pasteur-paris.rerif.ft.net (193.49.176.1) 3 ms 3 ms 3 ms 4 stlambert.rerif.ft.net (192.93.50.145) 7 ms 6 ms 6 ms 5 stamand3.renater.ft.net (195.220.180.9) 8 ms 8 ms 8 ms 6 stamand1.renater.ft.net (195.220.180.43) 7 ms 7 ms 7 ms 7 rbs1.renater.ft.net (195.220.180.50) 10 ms 7 ms 8 ms 8 Paris-EBS2.Ebone.net (192.121.156.89) 9 ms 9 ms 12 ms 9 icm-dc-2-S4/0-1984k.icp.net (192.157.65.129) 100 ms 105 ms 91 ms 10 icm-dc-1-F0/0.icp.net (198.67.131.36) 108 ms 97 ms 94 ms 11 icm-mae-e-H1/0-T3.icp.net (198.67.131.9) 303 ms 102 ms 298 ms 12 br2.tco1.alter.net (192.41.177.249) 94 ms 98 ms 99 ms 13 Hssi1-0.CR2.DCA1.Alter.Net (137.39.100.22) 100 ms 121 ms 122 ms 14 Fddi0-0.GW2.DCA1.Alter.Net (137.39.33.132) 92 ms 104 ms 123 ms 15 INSnet-gw1.ALTER.NET (137.39.128.130) 101 ms 94 ms 92 ms 16 s1.lon1gw1.uk.insnet.net (194.177.171.22) 176 ms 172 ms 188 ms 17 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 191 ms 187 ms 182 ms 18 collector.linx.net (194.68.130.254) 104 ms 100 ms * So I would really prefer a new root server near Washington. And I'm sure that would also be better to US biz'ness. Free suggestion to european network providers (Ebone and Dante particularly as those 2 are getting a lot of our money, I mean public money) : save the money for a root server and spend it in decent connectivity. (end of stupid-end-user comments). -- Christophe Wolfhugel -+- SIS, Institut Pasteur, Paris Boulot : wolf at pasteur.fr, $HOME : wolf at schnok.fr.net From kissg at sztaki.hu Wed Jan 22 09:40:13 1997 From: kissg at sztaki.hu (Gabor Kiss) Date: Wed, 22 Jan 1997 09:40:13 +0100 Subject: 2nd Root Server in Europe Message-ID: <199701220840.dN18691@sztaki.hu> > Try collector.linx.net = 194.68.130.254 > > However, quite a few networks have offered to supply transit to > the new root name server, so if anything connectivity should be > better than what it is now. >From 192.84.225.1 (AS3338 - AS1955) % trace 194.68.130.254 traceroute to 194.68.130.254 (194.68.130.254), 30 hops max, 40 byte packets 1 vhb.iif.hu (192.84.225.61) 2 ms 2 ms 2 ms 2 vha.iif.hu (192.84.229.61) 5 ms 3 ms 4 ms 3 vh-if.mta.iif.hu (193.6.206.89) 11 ms 11 ms 5 ms 4 mta-if.mtc.iif.hu (193.6.206.101) 40 ms 19 ms 16 ms 5 Amsterdam1-s2-1.dante.net (194.41.0.105) 159 ms 168 ms 172 ms 6 * New-York2.dante.net (194.41.0.42) 195 ms 284 ms 7 * f3-1.t32-0.New-York.t3.ans.net (204.149.4.9) 270 ms 249 ms 8 * * h9-1.t56-1.Washington-DC.t3.ans.net (140.223.57.29) 266 ms 9 enss149.t3.ans.net (140.223.57.22) 298 ms 277 ms 293 ms 10 192.41.177.112 (192.41.177.112) 315 ms 380 ms 307 ms 11 s1.lon1gw1.uk.insnet.net (194.177.171.22) 861 ms 620 ms 473 ms 12 * * atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 356 ms 13 collector.linx.net (194.68.130.254) 800 ms 424 ms * Ooops!! DANTE has no better connectivity (via BT) to 194.68.130.0!? (At least now.) Gabor From schneider at switch.ch Wed Jan 22 09:41:57 1997 From: schneider at switch.ch (Marcel Schneider) Date: Wed, 22 Jan 1997 09:41:57 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Message from Petri Helenius of "Tue, 21 Jan 1997 20:45:27 +0200." <199701211845.UAA07430@silver.sms.fi> References: <199701211845.UAA07430@silver.sms.fi> Message-ID: <9701220842.AA19669@ncc.ripe.net> On Tuesday, 21 Jan 1997, Petri Helenius writes: > At 14:21 21.1.1997 +0100, Marcel Schneider wrote: >>IMHO: Your answer is insufficient if not entirely mistaken. The inhabitants >>of central and southern Europe have a right to the same kind of connectivity >>as the northern region is now trying to get hold of or already has done >>so. Make no sense to improve an already good situation. >> > HAVE THE RIGHT? Excuse me? Just go to your telco and order the pipes and > stop whining. If they don't, tell everybody that they aren't up to their > job. Tell your legistrators that they de-regulate so somebody that _IS_ up > to the job can come in and do it properly. Very good suggestion. Thank you _very_ much. > Pete Marcel From Piet.Beertema at cwi.nl Wed Jan 22 10:41:32 1997 From: Piet.Beertema at cwi.nl (Piet Beertema) Date: Wed, 22 Jan 1997 10:41:32 +0100 Subject: 2nd Root Server in Europe In-Reply-To: "Your message of Wed, 22 Jan 1997 08:11:05 +0100 (MET) " <199701220711.IAA27512@yseult.sis.pasteur.fr> Message-ID: <9701220941.AA24296=piet@kraai.cwi.nl> Try collector.linx.net = 194.68.130.254 Bingo ! Still the end user's point of view : .... 7 rbs1.renater.ft.net (195.220.180.50) 10 ms 7 ms 8 ms 8 Paris-EBS2.Ebone.net (192.121.156.89) 9 ms 9 ms 12 ms 9 icm-dc-2-S4/0-1984k.icp.net (192.157.65.129) 100 ms 105 ms 91 ms 10 icm-dc-1-F0/0.icp.net (198.67.131.36) 108 ms 97 ms 94 ms 11 icm-mae-e-H1/0-T3.icp.net (198.67.131.9) 303 ms 102 ms 298 ms 12 br2.tco1.alter.net (192.41.177.249) 94 ms 98 ms 99 ms 13 Hssi1-0.CR2.DCA1.Alter.Net (137.39.100.22) 100 ms 121 ms 122 ms 14 Fddi0-0.GW2.DCA1.Alter.Net (137.39.33.132) 92 ms 104 ms 123 ms 15 INSnet-gw1.ALTER.NET (137.39.128.130) 101 ms 94 ms 92 ms 16 s1.lon1gw1.uk.insnet.net (194.177.171.22) 176 ms 172 ms 188 ms 17 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 191 ms 187 ms 182 ms 18 collector.linx.net (194.68.130.254) 104 ms 100 ms * We're a multi-homed site. The difference in routing between our two providers is small, as shown by these traceroute's, but what it comes down it that they *both* go via the US! Piet ---------------------------------------------------------- 1) Via NLnet: 1 cwi-gw (192.16.184.32) 3 ms 3 ms 3 ms 2 Amsterdam.NL.NL.net (192.16.183.96) 8 ms 5 ms 4 ms 3 Amsterdam4.NL.NL.net (194.178.244.40) 16 ms 31 ms 9 ms 4 Amsterdam2.NL.EU.net (134.222.18.1) 31 ms 251 ms 31 ms 5 * Vienna2.VA.US.EU.net (134.222.228.14) 190 ms 294 ms 6 maeegw1.us.insnet.net (192.41.177.112) 375 ms 302 ms 169 ms 7 s1.lon1gw1.uk.insnet.net (194.177.171.22) 220 ms 208 ms 214 ms 8 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 192 ms 189 ms 193 ms 9 collector.linx.net (194.68.130.254) 122 ms * 127 ms 2) Via SURFnet: 1 cwi-gw (192.16.184.32) 3 ms 4 ms 2 ms 2 Amsterdam11.router.surfnet.nl (192.16.183.112) 5 ms 5 ms 5 ms 3 Amsterdam9.router.surfnet.nl (145.41.6.67) 7 ms 5 ms 9 ms 4 New-York2.dante.net (145.41.0.14) 236 ms New-York2.dante.net (145.41.0.2) 196 ms New-York2.dante.net (145.41.0.14) 146 ms 5 f3-1.t32-0.New-York.t3.ans.net (204.149.4.9) 203 ms 209 ms * 6 h9-1.t56-1.Washington-DC.t3.ans.net (140.223.57.29) 240 ms 176 ms 195 ms 7 maeegw1.us.insnet.net (192.41.177.112) 244 ms enss149.t3.ans.net (140.223.57.22) 268 ms 264 ms 8 maeegw1.us.insnet.net (192.41.177.112) 226 ms 193 ms 189 ms 9 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 351 ms s1.lon1gw1.uk.insnet.net (194.177.171.22) 349 ms 315 ms 10 collector.linx.net (194.68.130.254) 124 ms * * From Ian.Mason at zoo.co.uk Wed Jan 22 10:45:44 1997 From: Ian.Mason at zoo.co.uk (Ian Mason) Date: Wed, 22 Jan 1997 09:45:44 +0000 Subject: 2nd Root Server in Europe (fwd) Message-ID: <3.0.32.19970120214637.006ca4ec@mail.zoo.net.uk.> At 13:08 20/01/97 +0000, Keith N Mitchell wrote: [snip] > I don't think there are any sacred cows for any of these here - > if we adjust the LINX policy to meet requirements of the root NS > rather than the other way around, I don't think we break either > of 1 or 2. I don't think it breaks 3. either, but this requirement > will go away when we have a route server in any case. If we opt for the root NS to have its own AS then there is no barrier to people offering identical routing policy for the LINX and the NS. If it shares the LINX AS then it will be impossible to have different policy for LINX and the NS. So, if there is a possibility that there will ever be a requirement to have different routing policy for the NS and the LINX it would seem wise to give the NS its own AS. It may be moot in as much as it's probably desireable to achieve 'root server connectivity quality' for the LINX AS, but the seperate AS for the root server would preserve some flexibility. From Piet.Beertema at cwi.nl Wed Jan 22 10:53:27 1997 From: Piet.Beertema at cwi.nl (Piet Beertema) Date: Wed, 22 Jan 1997 10:53:27 +0100 Subject: 2nd Root Server in Europe In-Reply-To: "Your message of Tue, 21 Jan 1997 15:59:54 +0100 (MET) " <199701211459.PAA17483@yseult.sis.pasteur.fr> Message-ID: <9701220953.AA24342=piet@kraai.cwi.nl> The location of that server is not a technical issue, it's a political one. True. And it will stay that way on this continent, where for ages and ages already national issues have dominated what politicians have the guts to call "European politics". There is no such thing: "European politics" is the sum of national politics and always the lowest common denominator. There's more than enough proof of that: "Bosnia" is one of the most striking proofs of the total failure of "European politics" (and since you're from France and I'm from Holland, I would add "drugs" to that... ;-)). In networking land it's no better. And there it's even worse, since the problem is made even worse by the price policies of the various national PTT's; deregulation, even though it has been the focus of the European Commission for many years, is largely still a dead letter. Just have a look at the prices that the various PTT's charge for international lines, and compare them with the prices for intercontinental lines. Then it isn't really surprising that in networking context the US has become the "center of Europe".... :-( Piet From jdd at vbc.net Wed Jan 22 11:22:49 1997 From: jdd at vbc.net (Jim Dixon) Date: Wed, 22 Jan 1997 10:22:49 +0000 (GMT) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <3.0.32.19970120214637.006ca4ec@mail.zoo.net.uk.> Message-ID: On Wed, 22 Jan 1997, Ian Mason wrote: > At 13:08 20/01/97 +0000, Keith N Mitchell wrote: > > I don't think there are any sacred cows for any of these here - > > if we adjust the LINX policy to meet requirements of the root NS > > rather than the other way around, I don't think we break either > > of 1 or 2. I don't think it breaks 3. either, but this requirement > > will go away when we have a route server in any case. > > If we opt for the root NS to have its own AS then there is no barrier to > people offering identical routing policy for the LINX and the NS. If it > shares the LINX AS then it will be impossible to have different policy for > LINX and the NS. Why is that? Different prefixes in the same AS can be treated differently. > So, if there is a possibility that there will ever be a > requirement to have different routing policy for the NS and the LINX it > would seem wise to give the NS its own AS. -- Jim Dixon VBCnet GB Ltd http://www.vbc.net tel +44 117 929 1316 fax +44 117 927 2015 From liman at sunet.se Wed Jan 22 12:28:21 1997 From: liman at sunet.se (Lars-Johan Liman) Date: Wed, 22 Jan 1997 12:28:21 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Your message of "Tue, 21 Jan 1997 19:01:40 +0000 (GMT)" References: Message-ID: <199701221128.MAA00290@fliptop.liman.sunet.se> On Tue, 21 Jan 1997, Gabor Kiss wrote: >> If anyone tell an address form the network >> what is the planned location of the new root server, >> all of us could test the good connectivity. >> E.g. with traceroute. jdd at vbc.net: > Try collector.linx.net = 194.68.130.254 Wasn't the idea to put it on a new prefix on a special port in a possibly new AS? A bit difficult to check beforehand in that case, right? Cheers, /Liman #------------------------------------------------------------------------- # Lars-Johan Liman ! Internet: liman at sunet.se # Ebone/NORDUnet/SUNET Operations Centre ! BITNET : LIMAN at SEARN # Royal Institute of Technology, Sweden ! HTTP : //www.sunet.se/~liman # ! Voice : Int +46 8 - 790 65 60 #------------------------------------------------------------------------- From GeertJan.deGroot at ripe.net Wed Jan 22 12:34:20 1997 From: GeertJan.deGroot at ripe.net (Geert Jan de Groot) Date: Wed, 22 Jan 1997 12:34:20 +0100 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of "Wed, 22 Jan 1997 10:22:49 GMT." Message-ID: <9701221134.AA24196@ncc.ripe.net> On Wed, 22 Jan 1997 10:22:49 +0000 (GMT) Jim Dixon wrote: > > If we opt for the root NS to have its own AS then there is no barrier to > > people offering identical routing policy for the LINX and the NS. If it > > shares the LINX AS then it will be impossible to have different policy for > > LINX and the NS. > Why is that? Different prefixes in the same AS can be treated differently. Please don't. If the routing policy is different, then the AS should be different. I don't think it's wise to do otherwise with critical resources like root nameservers. Geert Jan From liman at sunet.se Wed Jan 22 12:44:31 1997 From: liman at sunet.se (Lars-Johan Liman) Date: Wed, 22 Jan 1997 12:44:31 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Your message of "Wed, 22 Jan 1997 08:11:05 +0100 (MET)" References: <199701220711.IAA27512@yseult.sis.pasteur.fr> Message-ID: <199701221144.MAA00411@fliptop.liman.sunet.se> wolf at pasteur.fr: > Free suggestion to european network providers (Ebone and Dante particularly > as those 2 are getting a lot of our money, I mean public money) : save the > money for a root server and spend it in decent connectivity. How do you mean that Ebone gets a lot of "your" (public) money? /Liman #------------------------------------------------------------------------- # Lars-Johan Liman ! Internet: liman at sunet.se # Ebone/NORDUnet/SUNET Operations Centre ! BITNET : LIMAN at SEARN # Royal Institute of Technology, Sweden ! HTTP : //www.sunet.se/~liman # ! Voice : Int +46 8 - 790 65 60 #------------------------------------------------------------------------- From kissg at sztaki.hu Wed Jan 22 13:17:09 1997 From: kissg at sztaki.hu (Gabor Kiss) Date: Wed, 22 Jan 1997 13:17:09 +0100 Subject: 2nd Root Server in Europe Message-ID: <199701221217.KF21569@sztaki.hu> > The location of that server is not a technical issue, it's a political one. > True. And it will stay that way on this continent, where for > ages and ages already national issues have dominated what Sorry, I don't understand something. Why cannot a region set up a root name server for its private use? Users who can reach it fast are advised to put the address of this server into named cache file, others not. Generally peoples might select a appropriate set of root name servers and put them into cache file. This might be the menu: "Veeery official" servers: a.root-servers.net ... i.root-servers.net "Official" continental servers: a.europe.root-servers.net b.europe.root-servers.net k.europe.root-servers.net q.asia.root-servers.net "Semi-official" regional servers: s.north.europe.root-servers.net n.south.europe.root-servers.net "International ISP" servers: provider1.west.europe.root-servers.net provider2.west.europe.root-servers.net etc. There would be suggested root NS sets tailored for each region. Asian resolvers probably are not interested in a European root NS and vice versa. The only thing to discuss whether Internic does allow to download the zonefile for these hosts. Semi-official and International ISP servers may be tertiary root name servers even. OK, this idea may be junk. I just wished to start a brainstorming. Keep flame on minimum. Thanks for your patience. Gabor ------- Another funny idea, related to this one: The DNS itself could store a pointer to a suggested cache file specific to the requestor's address. (And a version number too.) E.g. cache.arpa. TXT "14,ftp://ftp.internic.net/pub/SOME_FILE" 193.cache.arpa. TXT "71,ftp://ftp.ripe.net/root-ns/europe.cache" 6.193.cache.arpa. TXT "4,ftp://ftp.hungarnet.hu/pub/hungarnet.cache" Users are suggested to find the longest name matching your address, and use cache file pointed by it. From tonyb at pipex.net Wed Jan 22 14:00:30 1997 From: tonyb at pipex.net (Tony Barber) Date: Wed, 22 Jan 1997 13:00:30 +0000 (GMT) Subject: 2nd Root Server in Europe In-Reply-To: <9701220941.AA24296=piet@kraai.cwi.nl> from "Piet Beertema" at Jan 22, 97 10:41:32 am Message-ID: <19970122130030.6938.qmail@pool.pipex.net> Piet Beertema wrote: > > Try collector.linx.net = 194.68.130.254 > Bingo ! > Still the end user's point of view : > .... > 7 rbs1.renater.ft.net (195.220.180.50) 10 ms 7 ms 8 ms > 8 Paris-EBS2.Ebone.net (192.121.156.89) 9 ms 9 ms 12 ms > 9 icm-dc-2-S4/0-1984k.icp.net (192.157.65.129) 100 ms 105 ms 91 ms > 10 icm-dc-1-F0/0.icp.net (198.67.131.36) 108 ms 97 ms 94 ms > 11 icm-mae-e-H1/0-T3.icp.net (198.67.131.9) 303 ms 102 ms 298 ms > 12 br2.tco1.alter.net (192.41.177.249) 94 ms 98 ms 99 ms > 13 Hssi1-0.CR2.DCA1.Alter.Net (137.39.100.22) 100 ms 121 ms 122 ms > 14 Fddi0-0.GW2.DCA1.Alter.Net (137.39.33.132) 92 ms 104 ms 123 ms > 15 INSnet-gw1.ALTER.NET (137.39.128.130) 101 ms 94 ms 92 ms > 16 s1.lon1gw1.uk.insnet.net (194.177.171.22) 176 ms 172 ms 188 ms > 17 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 191 ms 187 ms 182 ms > 18 collector.linx.net (194.68.130.254) 104 ms 100 ms * > >We're a multi-homed site. The difference in routing between >our two providers is small, as shown by these traceroute's, >but what it comes down it that they *both* go via the US! > > > Piet > >---------------------------------------------------------- > >1) Via NLnet: > 1 cwi-gw (192.16.184.32) 3 ms 3 ms 3 ms > 2 Amsterdam.NL.NL.net (192.16.183.96) 8 ms 5 ms 4 ms > 3 Amsterdam4.NL.NL.net (194.178.244.40) 16 ms 31 ms 9 ms > 4 Amsterdam2.NL.EU.net (134.222.18.1) 31 ms 251 ms 31 ms > 5 * Vienna2.VA.US.EU.net (134.222.228.14) 190 ms 294 ms > 6 maeegw1.us.insnet.net (192.41.177.112) 375 ms 302 ms 169 ms > 7 s1.lon1gw1.uk.insnet.net (194.177.171.22) 220 ms 208 ms 214 ms > 8 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 192 ms 189 ms 193 ms > 9 collector.linx.net (194.68.130.254) 122 ms * 127 ms > >2) Via SURFnet: > 1 cwi-gw (192.16.184.32) 3 ms 4 ms 2 ms > 2 Amsterdam11.router.surfnet.nl (192.16.183.112) 5 ms 5 ms 5 ms > 3 Amsterdam9.router.surfnet.nl (145.41.6.67) 7 ms 5 ms 9 ms > 4 New-York2.dante.net (145.41.0.14) 236 ms New-York2.dante.net (145.41.0.2) 196 ms New-York2.dante.net (145.41.0.14) 146 ms > 5 f3-1.t32-0.New-York.t3.ans.net (204.149.4.9) 203 ms 209 ms * > 6 h9-1.t56-1.Washington-DC.t3.ans.net (140.223.57.29) 240 ms 176 ms 195 ms > 7 maeegw1.us.insnet.net (192.41.177.112) 244 ms enss149.t3.ans.net (140.223.57.22) 268 ms 264 ms > 8 maeegw1.us.insnet.net (192.41.177.112) 226 ms 193 ms 189 ms > 9 atm0-2.lon2gw2.uk.insnet.net (194.177.170.196) 351 ms s1.lon1gw1.uk.insnet.net (194.177.171.22) 349 ms 315 ms >10 collector.linx.net (194.68.130.254) 124 ms * * > Folks, the reason for the long path via the US is partly our fault. Sorry. A quick fix on our filters should soon see connections via EBONE and PIPEX directly. Thanks for bringing this to our attention. Regards --Tony From knm at dial.pipex.com Wed Jan 22 12:18:02 1997 From: knm at dial.pipex.com (Keith N Mitchell) Date: Wed, 22 Jan 1997 11:18:02 +0000 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Ian Mason "Re: 2nd Root Server in Europe (fwd)" (Jan 22, 9:45am) References: <3.0.32.19970120214637.006ca4ec@mail.zoo.net.uk.> Message-ID: <970122153059.ZM7287@boiler.linx.org> On Jan 22, 9:45am, Ian Mason wrote: > > I don't think there are any sacred cows for any of these here - > > if we adjust the LINX policy to meet requirements of the root NS > > rather than the other way around, I don't think we break either > > of 1 or 2. I don't think it breaks 3. either, but this requirement > > will go away when we have a route server in any case. > > If we opt for the root NS to have its own AS then there is no barrier to > people offering identical routing policy for the LINX and the NS. If it > shares the LINX AS then it will be impossible to have different policy for > LINX and the NS. So, if there is a possibility that there will ever be a > requirement to have different routing policy for the NS and the LINX it > would seem wise to give the NS its own AS. > > It may be moot in as much as it's probably desireable to achieve 'root > server connectivity quality' for the LINX AS, but the seperate AS for the > root server would preserve some flexibility. I agree, but at the cost of an additional router. Due to the nature of the full routing we need for theses AS(es), a 64Mb router is a requirement eventually anyway, but it is not a LINX budget item at present, and while there have been lots of offers of server hardware for the root NS, there have been none to date of a router. I suggest we make the ASes the same for now, but split them when a clear requirement arises. At that point we will hopefully have a route server box to take over various of the collector functions. Keith From Daniel.Karrenberg at ripe.net Wed Jan 22 16:37:44 1997 From: Daniel.Karrenberg at ripe.net (Daniel Karrenberg) Date: Wed, 22 Jan 1997 16:37:44 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Your message of Wed, 22 Jan 1997 10:53:27 +0100. <9701220953.AA24342=piet@kraai.cwi.nl> References: <9701220953.AA24342=piet@kraai.cwi.nl> Message-ID: <9701221537.AA03077@ncc.ripe.net> > Piet Beertema writes: > The location of that server is not a technical issue, it's a political > one. > True. And it will stay that way on this continent, .... The RIPE DNS WG met and there was consensus that root servers should be connected through a dedicated address prefix and AS. Connectivity of the KTH based server will be re-engineered like this. The LINX based server will be set up like this. Further there was consensus that we should move *towards* a situation where the RIPE NCC is responsible for operating all root servers in the European region under the guidance RIPE and funded as an NCC core activity. It fits the principles for such activities extremely well (see below). It is important to note that while the NCC will have the operational responsibility this does not exclude interested parties to contribute to nameserver operation in any way. To the contrary the efforts of individuals and organisations like KTH, NORDUNET, LINX and VBCnet are very welcome and essential for reliable operation of the root nameservers. Given those two design decisions, root servers are not so closely bound to their location and can be moved around the network topology with relative ease as topology changes and other engineering parameters suggest. Although I am alightly biased in the RIPE NCC matter ;-) I think that this consensus both very sensible and quite solid. Daniel Principles for NCC Activities The RIPE NCC performs activities for the benefit of the Internet service provider s (ISPs) in Europe and the surrounding areas; primarily activities that the ISPs need to organise as a group, although they may be competing with each other in other areas. The RIPE NCC must therefore observe strict neutrality and impartiality with respect to individual service providers. In particular it refrains from activities that are clearly in the domain of the ISPs themselves. Activities are defined, performed, discussed and evaluated in an open manner. Results of activities such as software tools are made available to the public. Budgets as well as actual income and expenditure are published. Individual data will be kept in confidence where required. For example the amounts of address space allocated and assigned are published as are database entries of individual assignments including the relevant contact data; however the information supporting individual assignment requests is kept in strict confidence. While an activity may result in services being provided to an individual ISP, performing the activity as a whole must benefit the European ISPs as a group. For example address space registration services are provided to ISPs individually, but the activity as such benefits all ISPs by distributing address space according to common standards as well as maintaining a neutral and accessible registry. From pete at sms.fi Wed Jan 22 19:37:13 1997 From: pete at sms.fi (Petri Helenius) Date: Wed, 22 Jan 1997 20:37:13 +0200 (EET) Subject: 2nd Root Server in Europe Message-ID: <199701221837.UAA09916@silver.sms.fi> At 12:44 22.1.1997 +0100, Lars-Johan Liman wrote: >wolf at pasteur.fr: >> Free suggestion to european network providers (Ebone and Dante particularly >> as those 2 are getting a lot of our money, I mean public money) : save the >> money for a root server and spend it in decent connectivity. > >How do you mean that Ebone gets a lot of "your" (public) money? > Is this politics-wg at ripe.net :-) To be frank, it's impossible to follow the routes money takes around the European Union but as long as the speaker counts as a citizen of the union, there is at least some fraction of the money that he/she is paying in taxes. This specially count for the countries that pay more to the EU than receive from there, where I think France is one of them. (not discounting others, like Germany, Finland, Sweden, etc...) Pete From Piet.Beertema at cwi.nl Thu Jan 23 10:22:53 1997 From: Piet.Beertema at cwi.nl (Piet Beertema) Date: Thu, 23 Jan 1997 10:22:53 +0100 Subject: 2nd Root Server in Europe In-Reply-To: "Your message of Wed, 22 Jan 1997 16:37:44 +0100 " <9701221537.AA03077@ncc.ripe.net> Message-ID: <9701230922.AA27545=piet@kraai.cwi.nl> Given those two design decisions, root servers are not so closely bound to their location and can be moved around the network topology with relative ease as topology changes and other engineering parameters suggest. Then how about a mobile root server, moving from country to country, say, every month? Politically perfect! :-) Piet From Daniel.Karrenberg at ripe.net Thu Jan 23 10:43:03 1997 From: Daniel.Karrenberg at ripe.net (Daniel Karrenberg) Date: Thu, 23 Jan 1997 10:43:03 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Your message of Thu, 23 Jan 1997 10:22:53 +0100. <9701230922.AA27545=piet@kraai.cwi.nl> References: <9701230922.AA27545=piet@kraai.cwi.nl> Message-ID: <9701230943.AA19418@ncc.ripe.net> > Piet Beertema writes: > Given those two design decisions, root servers are not so > closely bound to their location and can be moved around the > network topology with relative ease as topology changes and > other engineering parameters suggest. > Then how about a mobile root server, moving from country > to country, say, every month? Politically perfect! :-) But not very practical or pragmatic. And that after all are our maximes. Daniel From robert.shaw at itu.int Thu Jan 23 13:18:57 1997 From: robert.shaw at itu.int (Robert Shaw) Date: Thu, 23 Jan 1997 13:18:57 +0100 Subject: proposed agenda - DNS WG session at RIPE26 References: <14718.853377004@kronos.NIC.DTAG.DE> Message-ID: <32E75731.3646@itu.int> Ruediger Volk wrote: > > please find below a proposed agenda for a DNS WG session at the > upcoming RIPE26. > [snip] > > 4. Report from IAHC proposal debate (mailing list) > > 5. Collect & prepare RIPE feedback to IAHC > Any quick results from this session? Cheers, Bob -- Robert Shaw (shaw at itu.int) Advisor, Global Information Infrastructure International Telecommunication Union (http://www.itu.int) Place des Nations, 1211 Geneva 20, Switzerland Note: IAHC member, expressing personal opinions From bmanning at ISI.EDU Fri Jan 24 10:49:32 1997 From: bmanning at ISI.EDU (Bill Manning) Date: Fri, 24 Jan 1997 01:49:32 -0800 (PST) Subject: 2nd Root Server in Europe (fwd) In-Reply-To: <9701221134.AA24196@ncc.ripe.net> from "Geert Jan de Groot" at Jan 22, 97 12:34:20 pm Message-ID: <199701240949.AA12127@zephyr.isi.edu> > > On Wed, 22 Jan 1997 10:22:49 +0000 (GMT) Jim Dixon wrote: > > > If we opt for the root NS to have its own AS then there is no barrier to > > > people offering identical routing policy for the LINX and the NS. If it > > > shares the LINX AS then it will be impossible to have different policy for > > > LINX and the NS. > > Why is that? Different prefixes in the same AS can be treated differently. > > Please don't. If the routing policy is different, then the AS should > be different. I don't think it's wise to do otherwise with critical > resources like root nameservers. > I would note that if there is a specific prefix and AS, then the router to support such an attachment need not be very large. This model is currently used for route servers and therouter does not have to carry full routing. -- --bill From GeertJan.deGroot at ripe.net Fri Jan 24 12:33:08 1997 From: GeertJan.deGroot at ripe.net (Geert Jan de Groot) Date: Fri, 24 Jan 1997 12:33:08 +0100 Subject: 2nd Root Server in Europe (fwd) In-Reply-To: Your message of "Fri, 24 Jan 1997 01:49:32 PST." <199701240949.AA12127@zephyr.isi.edu> Message-ID: <9701241133.AA17758@ncc.ripe.net> On Fri, 24 Jan 1997 01:49:32 -0800 (PST) Bill Manning wrote: > I would note that if there is a specific prefix and AS, then > the router to support such an attachment need not be very large. > This model is currently used for route servers and therouter > does not have to carry full routing. I don't think that is correct. It is likely that the router for the root-server (glad I don't need to pronounce that ;-) will get transit from multiple parties and need to do full routing as I don't think it is politically correct to default to someone. A route-server doesn't neccessary need to have full connectivity, and if it has, that connectivity doesn't have to be correctly spread among the people using it. That wouldn't be true for a root-server, would it? Geert Jan From GeertJan.deGroot at ripe.net Fri Jan 24 23:39:21 1997 From: GeertJan.deGroot at ripe.net (Geert Jan de Groot) Date: Fri, 24 Jan 1997 23:39:21 +0100 Subject: 2nd Root Server in Europe In-Reply-To: Your message of "Tue, 21 Jan 1997 16:56:58 +0100." <9701211557.AA04082@ncc.ripe.net> Message-ID: <199701242239.XAA00737@berklix.ripe.net> (this is slightly off-topic for dns-wg, but I hope it helps to make the issues at hand easier to understand. I don't think that dns-wg should become a discussion list for ISP economics, so please be considerate if you respond to this..) On Tue, 21 Jan 1997 16:56:58 +0100 Marcel Schneider wrote: > > As long as European major connectivity providers are ignoring each other > > and having their best interconnection points being MAE-East or the NY Nap > > all this seems useless to me... > > Agreements, policies. Also my (personal) opinion. But as long > as Peter Lothberg et al are attempting to tell us > 'route your inter-European IP packets over the US' nothing will improve. > The owners of 150+ MB links to the US can easily sell this opinion, > but it is absurd. Several people have told me they pay as much for a line to the USA, as a line between two European countries. The former is usually more useful, as for the latter you still need to obtain transit to other places, which has to be paid for too. Hence, a typical setup is a line to the USA, and some local peering, which is often only for national traffic. Personally, I think that for this reason Peter's comments are valid. It has nothing to do with Internet politics, but everything to do with extremely high prices for international lines in Europe. Europe hasn't changed much in this regard. Geert Jan