From s_wadge@cableinet.net Mon Jun 2 10:53:23 1997 From: s_wadge@cableinet.net (Steve Wadge) Date: Mon, 02 Jun 1997 10:53:23 +0100 Subject: 6Bone attachment. Message-ID: <33929813.59E2@cableinet.net> Dear All, I'm looking into getting CableInternet attached to the 6Bone, what is the procedure/steps for doing so? Do I need to register with RIPE first? Our likely attachment point is likely to be via UUNET. Regards, Steve Wadge, Data Network Engineer Tel: 01483 750900. From crawdad@fnal.gov Mon Jun 2 15:36:04 1997 From: crawdad@fnal.gov (Matt Crawford) Date: Mon, 02 Jun 1997 09:36:04 -0500 Subject: (IPng 3717) Re: proposal for RFC-1897 update In-Reply-To: "28 May 1997 13:57:53 +0200." <"970528135753.ZM2552"@rama.imag.fr> Message-ID: <199706021436.JAA19643@gungnir.fnal.gov> Alain says: > I've just read Bob proposal (draft-ietf-ipngwg-test-addr-alloc-01.txt). > Sounds like a very good start to me for the new addressing plan of the > 6-bone. If our registry can allocate NLAs (possibly to core 6bone sites), I > think we can move very quickly to this new addressing plan. I'd like to get an NLA for a Chicago-area research exchange called MREN, which is actually on the same switch as the Chicago NAP. I'm just back from vacation, but I think a TLA has been chosen for 6bone use already. Has someone taken on the delegation task? (Or is it being delegated?) Matt From guyd@uunet.pipex.com Mon Jun 2 19:14:30 1997 From: guyd@uunet.pipex.com (Guy Davies) Date: Mon, 2 Jun 1997 19:14:30 +0100 (BST) Subject: 6Bone attachment. In-Reply-To: <33929813.59E2@cableinet.net> Message-ID: Hi Steve, You're welcome to have a tunnel to us. If you use RIPng or static routing, we prefer that you connect solely via UUNET (you should see the mess multihoming with RIPng is causing). If you're using BGP4+, you're welcome to multihome :-) Contact ipv6@uunet.pipex.com for connection details. I'm out of the office tomorrow (Tuesday) so it may be Wednesday before I can get back to you. You can setup a RIPE entry once we're connected. Regards, Guy On Mon, 2 Jun 1997, Steve Wadge wrote: > Dear All, > > I'm looking into getting CableInternet attached to the 6Bone, what is > the procedure/steps for doing so? > Do I need to register with RIPE first? > Our likely attachment point is likely to be via UUNET. > > > Regards, > > Steve Wadge, > Data Network Engineer > Tel: 01483 750900. > From rzm@torun.pdi.net Tue Jun 3 01:09:46 1997 From: rzm@torun.pdi.net (Rafal Maszkowski) Date: Tue, 3 Jun 1997 02:09:46 +0200 (MET DST) Subject: 6BONE down? Message-ID: <199706030009.CAA17852@rymunda.torun.pdi.net> Having persisting problems with routing to NRL I tried to look around and hand-made the following table showing backbone sites reachability. 1st columnt contains a list of all backbone sites according to http://www-6bone.lbl.gov/6bone/6bone-bblinks.html 1st row lists the sites doing measurements, 2nd row - backbone sites they are connected to. Measurements are from different hours but more less Jun 2nd/3rd evening/night MET. 'x' means reachable '-' no measuments done The is a lot of sites completely down and some important ones appearing and disappering, like UUNET-UK (I couldn't get from PDi to UCAM-T today but now the route via UUNET-UK is back). Another site, important for a linuxer is NRL (and ftpV6 archive at Inner). It seems to be down completely. Because the table consist mostly of holes I'd like to start with e.g. NRL - has anybody seen a route to it? What are good methods to find problems in 6BONE? - I imagine that publishing raw routing table dump for all backbone sites, via ftp or WWW (or running public SNMP) would help a lot. R. #who JOIN NIST TU-BS CRS4 CSELT PDiT UCAM-T #via NRL,G6 JOIN CSELT? SICS IFB,UUNET-UK UNI-C x x x TELEBIT x x x x SICS x x x x x G6 x x x x x? JOIN x x x x WIDE x? SURFNET x ESNET x CICNET ISI-LAP NWNET x - x x VIAGENIE CISCO ANSNET IFB - x x NRL - CSELT x x ? UUNET-UK x DIGITAL-CA x x x x BAY x x x R. -- Rafal Maszkowski rzm@torun.pdi.net http://www.torun.pdi.net/~rzm Opinia publiczna powinna byc zaalarmowana swoim nieistnieniem - St. J. Lec From wessorh@ar.com Wed Jun 4 20:47:35 1997 From: wessorh@ar.com (Rick H. Wesson) Date: Wed, 4 Jun 1997 12:47:35 -0700 (PDT) Subject: need tunnle to connect to 6bone Message-ID: Hi, I have been tasked with connecting Organic Online (organic.com) to the 6bone. We are currently connected to the internet via UUnet. Could anyone provide us a tunnle? I can also be reached via rick@organic.com Thanks, -Rick =================================================================== Rick H. Wesson Internet Business Services rick@ar.com rom majordom@ISI.EDU Wed Jun 4 05:39:32 1997 Received: by zephyr.isi.edu (5.65c/5.61+local-26) id ; Wed, 4 Jun 1997 12:39:53 -0700 Received: from venera.isi.edu by zephyr.isi.edu (5.65c/5.61+local-26) id ; Wed, 4 Jun 1997 12:39:48 -0700 Received: from sgi.sgi.com by venera.isi.edu (5.65c/5.61+local-29) id ; Wed, 4 Jun 1997 12:39:39 -0700 Received: from newt.engr.sgi.com ([150.166.75.47]) by sgi.sgi.com (950413.SGI.8.6.12/970507) via ESMTP id MAA23044 for <@sgi.engr.sgi.com:6bone@ISI.EDU>; Wed, 4 Jun 1997 12:39:38 -0700 env-from (ms@newt.engr.sgi.com) Received: from newt.engr.sgi.com by newt.engr.sgi.com via ESMTP (950413.SGI.8.6.12/940406.SGI.AUTO) for <6bone@ISI.EDU> id MAA25063; Wed, 4 Jun 1997 12:39:32 -0700 Message-Id: <199706041939.MAA25063@newt.engr.sgi.com> X-Mailer: exmh version 1.6.5 12/8/95 To: 6bone@ISI.EDU Subject: attachment Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 04 Jun 1997 12:39:32 -0700 From: Mark Smith Sender: owner-6bone@ISI.EDU Precedence: bulk Hello, I think SGI is ready to look into attaching to 6bone. I guess Cisco (6bone-router.cisco.com) is the most likely attachment point. Please let me know if this is correct. A list of the procedure to attach would be helpful too. Thanks. From davidk@isi.edu Wed Jun 4 22:17:34 1997 From: davidk@isi.edu (davidk@isi.edu) Date: Wed, 4 Jun 1997 14:17:34 -0700 (PDT) Subject: new RIPE-style routing registry conversion Message-ID: <9706042117.AA27812@brind.isi.edu> Hi, I am ready to do the conversion to the new registry as already announced earlier by Bob. I would like to carry out this operation during this weekend if there are no objections. I will try to convert the data as best as possible using some smart scripts and manual editing but please realize that it is not always possible due to lack of consistency in the current data set and the already quite large number of objects. I would like to ask if everybody could check their data in the new registry after the conversion is completed and update any data that is incorrect. Don't hesitate to send me questions if you have problems doing so, but please with this until I send an announcement that the conversion has been completed. I will send out a mail after the conversion has been done with full details and pointers to information on how to work with the new registry. In the mean time, check out: http://www.isi.edu/~davidk/6bone/ for a preview of the new registry. Note that updates to the current test registry will stop working soon due to the switchover to our new dedicated machine (whois.6bone.net). David K. --- From thamdi@rocketmail.com Fri Jun 6 16:34:44 1997 From: thamdi@rocketmail.com (T. Hamdi) Date: Fri, 6 Jun 1997 08:34:44 -0700 (PDT) Subject: Beginner need help Message-ID: <199706061533.IAA12437@mta5.rocketmail.com> Hello, Can someone tell me wich steps to follow to hook to the 6bone ? i have a machine with and ipv6 kernel (solaris) wich asked me some adresses to start ipv6'ing. Thanks in advance. _____________________________________________________________________ Sent by RocketMail. Get your free e-mail at http://www.rocketmail.com From b82072@ccstudent.ee.ntu.edu.tw Sun Jun 8 09:28:41 1997 From: b82072@ccstudent.ee.ntu.edu.tw (Jeng-hung Tsai) Date: Sun, 08 Jun 1997 16:28:41 +0800 Subject: Looking for connection point Message-ID: <3.0.2.32.19970608162841.006a623c@cctwin.ee.ntu.edu.tw> Hello, I am looking for a connecting point to 6bone. I am in Taiwan. What is the most appropriate site for me? Jeng-hung Tsai, 6/8 From davidk@isi.edu Mon Jun 9 07:29:15 1997 From: davidk@isi.edu (davidk@isi.edu) Date: Sun, 8 Jun 1997 23:29:15 -0700 (PDT) Subject: New registry operational Message-ID: <9706090629.AA07393@brind.isi.edu> Hi all, As announced last week, I have now converted all 158 (!) 6BONE ftp registry files to the new registry format. Some parts of entries could not be converted correctly due to some inconstencies in the old data set. Please update these entries yourself if you find that to be the case. Check out http://www.6bone.net/RIPE-registry.html on how to update entries in a RIPE style database. For those of you that are already familiar with this type of databases, send your updates to auto-dbm@ISI.EDU. Most inconsitencies had to do with reverse DNS not working or missing protocol information in the tunnel fields. Please check my latest draft on how the entries should look like: http://www.isi.edu/~davidk/6bone/draft-ietf-ngtrans-6bone-registry-02.txt Feel free to contact me directly if you find inconsistencies that you cannot solve yourself or suspect a bug in this quite new piece of software. You can query the database at whois.6bone.net with any 'whois' tool or even 'telnet whois.6bone.net whois'. However, you can also download a special version of 'whois' for better results (with special 6BONE support): http://www.isi.edu/~davidk/6bone/ripe-whois-tools-3.0.0+6bone-extensions.tar.gz A simple webinterface is also available at: http://www.isi.edu/~davidk/6bone/whois.html The special whois client (including the webinterface) has a very nice new feature that can handle references to external objects. This means that it can now resolve RIPE or InterNIC NIC handle references which allows you to use your existing NIC handle in the 'contact:' field. (for an example: try a lookup for SWITCH) Another feature can assist you in finding a topological nearby ipv6-site: $ whois -h whois.6bone.net -i origin AS1835 will give you all ipv6-sites that have AS1835 as their origin AS. And $ whois -h whois.6bone.net -r -i tunnel unvea.denet.dk will give you all objects that have 'unvea.denet.dk' in their tunnel specification. And $ whois -h whois.6bone.net 5F07:2B00::/128 will give you the site that is possibly using this IPv6 address. A full dump of the database plus serial number is made available daily at: ftp://whois.6bone.net/6bone/6bone.db.gz The copyright message in the file is solely to scare spammers and other abusers. It can be used to do queries that are not supported by the whois interface or people that want to run a local mirror. We plan to have an official backup site at LBL and several mirrors around the world for which we already have several volunteers. We will inform you on the list when they come available. I will try to get the documentation up to date as soon as possible and will ask RIPE to close down the old FTP style registry. The data in that registry will still be available in case you need it: ftp://whois.6bone.net/6bone/old_RIPE_ftp_registry.tar.gz I hope this helps, David K. --- PS1 Two useful scripts that people asked for in the past PS2 RIPE database configuration for those of you that are already running the RIPE database software and want to configure it to run a mirror site --- #!/bin/sh # # this script finds the first host # in the 6bone registry that is providing # a specific service # # example usage: # # $ ping6 `applicationhost uunet-uk ping` whois -h 6bone -r $1 | \ awk '{if (($1 == "application:") && \ ($2 == "'$2'") && \ (!i)) { \ print $3; i=1 \ }}' # end of script --- #!/bin/sh # # host6 lists the DNS name and IPv4 & IPv6 address of a host # # example: host6 unvea.denet.dk # # It uses the 'host' tool by Eric Wassenaar # # get it from: ftp://ftp.nikhef.nl/pub/network/host.tar.Z # (the directory also includes some other very good # replacement of traceroute and other networking tools) # echo Name: $1 host -t A $1 2>&1 | awk '{ if ($3 ~ /[0-9]/) { print "IPv4 address:", $3 } }' host -t AAAA $1 2>&1 | awk '{ if ($3 ~ /\:/) { print "IPv6 address:", $3 } }' # end of script --- RIPE database configuration file: ATTR is ipv6-site CLASSLESS ATTR lo location ATTR pf prefix i6 ATTR ap application ATTR tl tunnel ATTR ct contact pn,ro ATTR ul url # ipv6-site # OBJ is ATSQ is or de lo cy pf ap tl ct rm ul ny mb ch so OBJ is MAND is or de cy pf ct ch so OBJ is MULT lo de cy pf ap tl ct rm ul ny mb ch OBJ is UNIQ is OBJ is KEYS is pf OBJ is REC ct pf --- From thamdi@rocketmail.com Mon Jun 9 14:03:12 1997 From: thamdi@rocketmail.com (T. Hamdi) Date: Mon, 9 Jun 1997 06:03:12 -0700 (PDT) Subject: TUNNEL NEEDED Message-ID: <199706091304.GAA00170@mta5.rocketmail.com> Hello, i think my last message was lost so here's another. All i need is someone setting a tunnel for me. we are have a sprint link to the usa. can someone help me please ? _____________________________________________________________________ Sent by RocketMail. Get your free e-mail at http://www.rocketmail.com From dth@lucent.com Mon Jun 9 14:45:44 1997 From: dth@lucent.com (Harrington, Dan) Date: Mon, 9 Jun 1997 09:45:44 -0400 Subject: Connecting to 6bone guidelines (*VERY* rough draft) Message-ID: Hello all, The enclosed is a preliminary (and incomplete) draft of a document which I had envisioned as a web page for www.6bone.net, as a sort of FAQ (in non-Q/A format). Bob Fink reviewed the outline, and had suggested that it might be a worthwhile I-D for the 6bone group as well. In any event, the flurry of mail messages from folks wanting to connect to the 6BONE made me decide to throw this out to the group now, incomplete or not, for feedback. What I would like to get from reviewers are: - low-level review for accuracy, clarity, grammar, etc. - suggestions for additional material, topics, expansion... - comments on applicability of this as an Internet-Draft Thanks, Dan ====================================================================== Q: Style question..."6BONE" or "6bone"? NB: Using new registry and addressing plan info, although neither is yet operational. (OK, David, you made a liar of me... :-) Introduction: This document is an attempt to provide guidance for those wishing to participate in the ongoing 6BONE experiment. The 6BONE is a virtual network of IPv6-capable systems, which utilizes the world-wide connectivity of the Internet to link researchers, developers and users of IPv6. The scope of work on this testbed is varied and variable, but the overriding focus is to develop protocols and applications related to the IPv6 deployment. For further details about the 6BONE, including how to join the 6BONE user mailing list, please see the official Web site at http://www.6bone.net/. In the following sections, we will review the steps required to connect to the 6BONE. These steps have been organized as a series of stages: Planning, Training, and Deployment. Please keep in mind that these are fairly arbitrary stages, meant mainly to organize the data in the document, but it may also help to organize you as you proceed. Finally, this document is only as useful as the information it contains. If you have any suggestions regarding content or style, and especially if you discover any errors, please contact the author with your comments. Thank you. Planning ======== The most important (and possibly the most time consuming) step you will take on the road to 6BONE connectivity is that of planning what you will do. If you organize yourself at the beginning of the task, you will find that the completion of the task is smooth and simple. So even if you have to get connected to the 6BONE this minute, and not a moment later, please take the time to review the following items; you just might save yourself some effort and frustration! The first thing to consider before connecting to the 6BONE is what you will do when you get there. The 6BONE is a test network, made up almost entirely of prototype software in various stages of development. It is not a final destination really, but a road towards a future Internet. And this particular road is not a superhighway, but more like the first road to a pioneer community, somewhat bumpy, full of adventure and peril, and populated by intrepid travelers. Before you start down this road you should form some idea of where you want to end up. The reason need not be deeply profound! You may be a network operator who plans to support this new family of protocols for your users or customers, and need to gain operational experience with the new addressing formats and new tools. Or you may be a researcher or developer working on a new protocol or application. You may merely be driven to use the newest technology, and to figure out new tools and techniques. In any event, welcome! It's not important what your purpose is on the 6BONE, but rather that you have one. The next issue to consider is what type of network you are planning to support. It may be a small set of systems under your direct control, or it could be a larger set of disperse systems within your larger network. The mechanisms you will use to coordinate activities in these distinct environments will be rather different. The amount and type of internal infrastructure you will require must take into consideration the types of users (naive or experienced), the systems and software you plan to deploy (homogeneous vs. completely mixed), and the management style used in your network (centrally controlled vs. distributed autonomy). In any event you should maintain mailing lists, telephone lists, and supported procedures for regularly scheduled events in your network. [This last bit doesn't seem clear...try to clarify, Dan] Finally, the type of 6BONE connectivity you plan will dictate the level of support you will dedicate to this activity. The 6BONE currently supports three levels of hierarchy, with leaf sites having or more connections into the 6BONE, purely for their own reachability. Transit sites, as their name indicates, permit data to travel through their site from leaf networks, in addition to terminating their own communications. The highest levels of support would be required for core sites, which maintain the 6BONE's backbone connectivity with a high proportion of other core sites in addition to providing access to one or more transit or leaf sites. This breakdown of the 6BONE hierarchy can be seen in the current 6BONE map, at http://www.6bone.net/6bone-drawing.html. Another important aspect of planning will be to determine what software you will be using in your network. There are a number of different packages available for evaluating IPv6, ranging from freely-available source code for popular UNIX implementations to commercially available kits for both hosts and routers. In order to determine what best fits your needs, a bit of research is in order. You can get a good idea of what is currently available by checking the "Implementations" page under the IPng homepage, at URL http::/playground.sun.com/ipng/ipng-implementations.2.html, where both host and router software is described, often with pointers to further information. If you know in advance which systems you will be using for IPv6 testing, you may wish to gather current version information, for quick cross-referencing against what is required. It is also important to note just which capabilities you will require (e.g. DNS/BIND server, Basic API support), as this may help to focus your search. The minimum set of capabilities required by most 6BONE sites would be a router (i.e., the ability to forward packets off-site to the 6BONE), a DNS/BIND server capable of storing AAAA records and resolving your branch of the ip6.int tree [although this service could potentially be provided by other means, e.g. by 6BONE transit site...could be clunky, but should this be in minimum set? dth], and a host with ping, nslookup and traceroute utilities. [Again, this last set is arbitrarily defined by me...anybody feel like arguing? dth]. Bear in mind that a single system is quite capable of performing all these functions, although it may be more typical to distribute this work among several machines. Having determined what you plan to do, and what systems you plan to do it with, the final and critical planning step remains; where will you connect to the 6BONE? Because the 6BONE is not a simple Internet Service Provider, but a loose agglomeration of interested parties, there is no support center or NOC (Network Operations Center) to call for help. It is up to you to find someone already connected to the 6BONE who would be willing to provide access to your site. This is not as daunting a task as it might seem at first, for the people and organizations which are already connected to the 6BONE are there because they want to make it work, and you will certainly find a willing partner. The research you should do in this area is to determine who is already connected, and which of these sites are near yours. [Is the following useful? It seems like we should make an effort to be efficient when connecting...dth] Aside: Note that use of the "near" above is deliberately ambiguous. While it is highly desirable to utilize a contact located geographically close to your site, this may in fact not make the most sense from a topological point of view. It is important to keep in mind that the 6BONE is overlayed upon the existing Internet infrastructure, largely using IPv4 tunnels from site to site. Thus, a single IPv6 router-to-router hop may require multiple IPv4 hops to complete. For example, two independent networks located just a few miles apart may utilize different ISPs, such that traffic between them may travel 15-20 hops on average. On the other hand, two sites located in different states or countries may use the same ISP, and may thus traverse significantly fewer hops. Given that your 6BONE contact site will be the first hop onto the 6BONE, it makes sense to reduce the overhead of connectivity as much as possible. If you identify multiple potential 6BONE contacts, it would be a worthwhile exercise to spend some time running traceroute to determine how reachable they are from your network. [Is this a potential tool requirement for the 6BONE? As the 6BONE gets larger its growing pains will reflect its lack of relationship with the underlying IPv4 network. Has the MBONE gone through this stage already? Do some of their documents cover this ground? Check it out...dth] Some tools which can you can use to figure out where you might be able to connect include the 6BONE map (at URL ), the ISI 6BONE registry (at URL ), and the 6BONE mailing list itself. [This last method is not the preferred mechanism, but it usually does prove effective if all else fails.] The information you are looking for begins with a contact name, and should include both e-mail and telephone addresses. Make sure to inquire of a potential site which style of communication they prefer, and make a note of it...not everyone enjoys receiving telephone calls. Should the potential site be located in a different timezone or another country, make a note of this with your contact information, so that you remember to call at the correct time. Once a 6bone contact has been identified and has agreed to serve as your transit or backbone service provider, you are ready to develop your addressing plan. While the initial stages of 6bone deployment took place using RFC 1897 , this addressing plan has been superseded for a couple of reasons. First of all, it used IPv4 addressing elements (including the Autonomous System number) of a site to generate the IPv6 address. Because the 6bone uses the IPv4 network as a virtual network, and not as a strict topology guide, the address prefixes defined in this manner did not aggregate; that is, each site had a unique prefix, and routing was done on an essentially flat space. This type of plan is known to not scale well. Secondly, the IPng Working Group has made some fundamental changes to the IPv6 Addressing Architecure , which has resulted in a new unicast address assignment plan and a new 6bone address allocation plan prefix. Those sites, leaf or transit, which connect to these TLA's receive a first level Next-Level Aggregator identifier. There is a tertiary level which uses a third field known as the Site-Level Aggregator, which provides a field for subnet assignments. Together these fields form a 64-bit route to a particular network/subnet. The other 64-bit value in an address is the interface identifier. The IPv6 addresses in use on the 6bone would then have the following format: 0x3FFE<64-bit interface ID> Please check the current Internet-Drafts in this area (as referenced above) for further details. In summary, your 6bone address will be based on (and provided by) your 6bone connection point. In the event that you plan to be a 6bone core (or backbone) site, please review the registry information at www.6bone.net [good punt, eh?]. [Further info here re. addressing plan...small site vs. large network, flat routed vs. subnets, etc.] At the same time that you develop your addressing plan it is suggested that you should arrange for the delegation of your portion of the ip6.int tree in the Domain Name System [RFC1034] , which is described in RFC 1886. * Get DNS delegation, create zone files (see current info under 6BONE page for current hints/tips) Training ======== This can be as limited as you personally feel necessary in order to begin deployment, but may be more extensive if you include the training you may provide to operators, users, customers and programmers within your network. Some initial sources of information include the following: www=> IPng page, IETF's working group I-D page, vendor sites and white papers * Current IPv6 RFC's and Internet-Drafts * Books on the subject * Mbone sessions (IETF working groups, etc.) (???) * mailing lists (6bone, personal, other) * Magazine and journal articless * seminars and presentations * share what you learn! Write up experiences, tips, present them to user's group or other forum, answer queries on mailing list, etc. Deployment ========== Actually doing something at last! Put your plans into action...if you've planned well, this should be smooth and relatively quick. - Gather systems, software, equipment as per planning - Do preliminary builds, installations. Test internal connectivity, gain experience with tools, initialize DNS infrastructure. - Connect as planned...tie in zones to DNS...test tunnel or link as appropriate - Register in new RIPE style registry db at ISI - Announce yourself on 6BONE mailing list References ========== [RFC1034] P. Mockapetris, "Domain names - concepts and facilities", RFC1034, 11/01/1987 [RFC1886] S. Thomson, C. Huitema, "DNS Extensions to support IP version 6", RFC 1886, December 1995. ...and many more... From bmanning@ISI.EDU Mon Jun 9 16:16:03 1997 From: bmanning@ISI.EDU (bmanning@ISI.EDU) Date: Mon, 9 Jun 1997 08:16:03 -0700 (PDT) Subject: Looking for connection point In-Reply-To: <3.0.2.32.19970608162841.006a623c@cctwin.ee.ntu.edu.tw> from "Jeng-hung Tsai" at Jun 8, 97 04:28:41 pm Message-ID: <199706091516.AA04743@zed.isi.edu> > > Hello, > > I am looking for a connecting point to 6bone. > If you have not found one yet, I'd be happy to provide you with an entry point. IPv4 endpoint: 198.32.146.11 Protocol: RIPng IPv6 prefix: 5FBC:1000::0/32 -- --bill From Alain.Durand@imag.fr Tue Jun 10 14:50:06 1997 From: Alain.Durand@imag.fr (Alain Durand) Date: Tue, 10 Jun 1997 15:50:06 +0200 Subject: new addressing plan In-Reply-To: davidk@ISI.EDU "Re: new addressing plan" (Jun 9, 10:56am) References: <9706091756.AA27913@brind.isi.edu> Message-ID: <970610155006.ZM22236@rama.imag.fr> On Jun 9, 10:56am, davidk@ISI.EDU wrote: > Subject: Re: new addressing plan > > I can start assigning prefixes anytime as long as people tell me how to > do it. We have Bob's draft ideas but nobody really commented so far ... I think the proposed algorithm is ok enough to start. There is an assumption that we can not implement this addressing plan till new application with 64 bit IDs are avalaible. I think we can still assign prefixes before that time and see what is going on. In the meantime, we 'just' have not to use the 16 extra id bits. What we can do is to assign 8bits NLA1 to core site asking for a core prefix. This is 255 core prefixes, more than enough. Those sites will then assign longer prefixes to leaf sites and transit sites. How do we assign those numbers? well the +1 algorithm sounds reasonable to me. I'm now ready to get a /24 prefix for G6. - Alain. From roque@cisco.com Tue Jun 10 17:34:34 1997 From: roque@cisco.com (Pedro Marques) Date: Tue, 10 Jun 1997 09:34:34 -0700 Subject: new addressing plan In-Reply-To: <970610155006.ZM22236@rama.imag.fr> References: <9706091756.AA27913@brind.isi.edu> <970610155006.ZM22236@rama.imag.fr> Message-ID: <199706101634.JAA14680@trix.cisco.com> >>>>> "Alain" == Alain Durand writes: Alain> On Jun 9, 10:56am, davidk@ISI.EDU wrote: >> Subject: Re: new addressing plan >> >> I can start assigning prefixes anytime as long as people tell >> me how to do it. We have Bob's draft ideas but nobody really >> commented so far ... Alain> I think the proposed algorithm is ok enough to start. Is this the general opinion ? Personaly i believe our present aproach is superior since it does not require any registry... And simply swaping 010 with 001 and moving the reserved fields around will give us the new address format from rfc 1897. present format: | 3 | 5 bits | 16 bits | 8 | 24 bits | 8 | 16 bits|48 bits| +---+----------+----------+---+------------+---+--------+-------+ | | |Autonomous| | IPv4 | | Subnet | Intf. | |010| 11111 | System |RES| Network |RES| | | | | | Number | | Address | | Address| ID | +---+----------+----------+---+------------+---+--------+-------+ my proposal: | 3 | 5 bits | 16 bits | 24 bits | 16 bits| 64 bits | +---+----------+----------+------------+--------+-----------+ | | |Autonomous| IPv4 | Subnet | | |001| 11111 | System | Network | | | | | | Number | Address | Address| EID | +---+----------+----------+------------+--------+-----------+ Pedro. From hamdi.tounsi@ati.tn Tue Jun 10 19:56:58 1997 From: hamdi.tounsi@ati.tn (Hamdi TOUNSI) Date: Tue, 10 Jun 1997 17:56:58 -0100 Subject: ipv6 Prefix and local adress Message-ID: <339DA37A.BBF@ati.tn> Hello, someone at isi gave me the prefix : 5fbc:1000::0/32 should this be the prefix of my local ipv6 adress too ? another question. Can someone explain me the following adresses : - my ipv6 adress - a link ipv6 adress (entry of tunnel) - a link ipv6 adress (end of tunnel) - the tunnel endpoint ipv6 adress Peace Hamdi From roque@cisco.com Tue Jun 10 19:10:26 1997 From: roque@cisco.com (Pedro Marques) Date: Tue, 10 Jun 1997 11:10:26 -0700 (PDT) Subject: new addressing plan In-Reply-To: <3.0.32.19970610135443.006b75f4@pobox> References: <3.0.32.19970610135443.006b75f4@pobox> Message-ID: <199706101810.LAA27889@pedrom-ultra.cisco.com> >>>>> "Dimitry" == Dimitry Haskin writes: Dimitry> At 09:34 AM 6/10/97 -0700, Pedro Marques wrote: >> Is this the general opinion ? >> >> Personaly i believe our present aproach is superior since it >> does not require any registry... >> Dimitry> It is not that we can't get away without registry at Dimitry> 6bone for some time. To be more specific, we don't need, with an AS based scheme, to have address delegation being made by a registry. A registry is useful and the RIPE registry as been a great aid to everyone. Dimitry> It is that we need to start testing Dimitry> addressing concepts to their full extend asap. I see no advantage in having a registry assigned prefix and a prefix build from a AS #... the addressing concept is exactly the same. Dimitry> This includes registry operations and address delegation which Dimitry> are important components in overall scheme. At the moment i believe that requiring address delegation will only make things harder for people to get into the 6bone... things are hard enought as it is for most people. Dimitry> The fact Dimitry> that there is no operational v6 registry has become Dimitry> detrimental to the v6 deployment effort I think it is a plus... Dimitry> since it Dimitry> contributes to the perception that v6 lacks even basic Dimitry> deployment logistics. A 6bone registry would not be able to delegate non testing allocation addresses. Also i do believe that this is not a real technical problem but incorrect perception by some people... Pedro. From dhaskin@baynetworks.com Tue Jun 10 18:54:44 1997 From: dhaskin@baynetworks.com (Dimitry Haskin) Date: Tue, 10 Jun 1997 13:54:44 -0400 Subject: new addressing plan Message-ID: <3.0.32.19970610135443.006b75f4@pobox> At 09:34 AM 6/10/97 -0700, Pedro Marques wrote: > >Is this the general opinion ? > >Personaly i believe our present aproach is superior since it does not >require any registry... > It is not that we can't get away without registry at 6bone for some time. It is that we need to start testing addressing concepts to their full extend asap. This includes registry operations and address delegation which are important components in overall scheme. The fact that there is no operational v6 registry has become detrimental to the v6 deployment effort since it contributes to the perception that v6 lacks even basic deployment logistics. Dimitry From pcurran@ticl.co.uk Tue Jun 10 20:46:22 1997 From: pcurran@ticl.co.uk (Peter Curran) Date: Tue, 10 Jun 1997 20:46:22 +0100 Subject: new addressing plan Message-ID: <199706101957.UAA23089@gate.ticl.co.uk> > > Is this the general opinion ? > > Personaly i believe our present aproach is superior since it does not > require any registry... > I do not think this is GOOD. The stated aim of the 6bone is to gain operational experience of v6 deployment. That must include exploring the requirements of the registry management, database objects, etc. I appreciate that the small scale of the 6bone cannot hope to emulate the size of the final Internet requirement, but the principles we develop will be invaluable input into the final registry management system. I think that the work that is now underway in handling the new site registry is going to be very useful on this basis and support the efforts to get this off the ground. Likewise, I think that we need to start exercising the new address format in as real a form as we can manage for the same reasons. My .2p Cheers Peter TICL From davidk@isi.edu Tue Jun 10 21:52:46 1997 From: davidk@isi.edu (davidk@isi.edu) Date: Tue, 10 Jun 1997 13:52:46 -0700 (PDT) Subject: new addressing plan In-Reply-To: <199706101634.JAA14680@trix.cisco.com> from "Pedro Marques" at Jun 10, 97 09:34:34 am Message-ID: <9706102052.AA04304@brind.isi.edu> Pedro, Pedro Marques writes: > > Personaly i believe our present aproach is superior since it does not > require any registry... I agree that using the AS in someway is a very attractive way to go. The nice thing is that you don't need any human judgement whether you can get a prefix or not. However, the real world is not as simple as that. In the end the decision has to be made who gets and AS and who not, which will become increasingly difficult because we are giving more value to the concept of an 'AS' then it had in the past. Furthermore, it is already very common that an AS gets split up in several IPv6 spaces to accomodate customers of the provider and other people to join the 6bone. Doing this means that we need a (local) registry of some kind and it makes sense to provide such a service for everybody at once instead of letting all people to run their own database (I guess that most people on the 6bone can spend their time better). Furthermore, we are supposed to test things like aggregration and using an AS based structure will brake that from the beginning. Note that assigning the prefixes should not become a difficult thing. I don't see much reason why this could not be done in a highly automated fashion since we are dealing with temporary test addresses which are not very attractive to hoard. We might even be able to use the registry for renumbering purposes as a kind of global DCHP route slot server ... David K. --- From crawdad@fnal.gov Tue Jun 10 22:59:21 1997 From: crawdad@fnal.gov (Matt Crawford) Date: Tue, 10 Jun 1997 16:59:21 -0500 Subject: new addressing plan In-Reply-To: "10 Jun 1997 09:34:34 PDT." <"199706101634.JAA14680"@trix.cisco.com> Message-ID: <199706102159.QAA15225@gungnir.fnal.gov> > And simply swaping 010 with 001 and moving the reserved fields around will > give us the new address format from rfc 1897. > > my proposal: > > | 3 | 5 bits | 16 bits | 24 bits | 16 bits| 64 bits | > +---+----------+----------+------------+--------+-----------+ > | | |Autonomous| IPv4 | Subnet | | > |001| 11111 | System | Network | | | > | | | Number | Address | Address| EID | > +---+----------+----------+------------+--------+-----------+ This isn't consistent with draft-ietf-ipngwg-unicast-aggr-00.txt. Did you mean it to be? The above format would put people randomly into the same TLA if the first 8 bits of their provider's ASN matched. From roque@cisco.com Tue Jun 10 23:22:09 1997 From: roque@cisco.com (Pedro Marques) Date: Tue, 10 Jun 1997 15:22:09 -0700 (PDT) Subject: new addressing plan In-Reply-To: <199706102159.QAA15225@gungnir.fnal.gov> References: <"199706101634.JAA14680"@trix.cisco.com> <199706102159.QAA15225@gungnir.fnal.gov> Message-ID: <199706102222.PAA28016@pedrom-ultra.cisco.com> >>>>> "Matt" == Matt Crawford writes: >> And simply swaping 010 with 001 and moving the reserved fields >> around will give us the new address format from rfc 1897. >> >> my proposal: >> >> | 3 | 5 bits | 16 bits | 24 bits | 16 bits| 64 bits | >> +---+----------+----------+------------+--------+-----------+ >> | | |Autonomous| IPv4 | Subnet | | >> |001| 11111 | System | Network | | | >> | | | Number | Address | Address| EID | >> +---+----------+----------+------------+--------+-----------+ Matt> This isn't consistent with Matt> draft-ietf-ipngwg-unicast-aggr-00.txt. That depends if you consider that the bit where a boundary lies is significant. I believe the above is pretty much consistent with the idea of agregation by provider or large struture. But it was just a straw-man proposal... feel free to move the bits around. But anyway from the sample on the mailing list it seams people prefer a registry delegation based scheme to an automatic one... Pedro. From AConta@lucent.com Wed Jun 11 00:24:07 1997 From: AConta@lucent.com (Conta, Alex) Date: Tue, 10 Jun 1997 19:24:07 -0400 Subject: ipv6 Prefix and local adress Message-ID: > Hello, > someone at isi gave me the prefix : 5fbc:1000::0/32 > should this be the prefix of my local ipv6 adress too ? > No the prefix of your link local address is a predefined hardcoded value: FFFE:8000::0/18 This looks like a global prefix so this should be used to build your interface's global IPv6 address - append the MAC address to it, for instance for a 0a:0b:0c:0d:0e:0f MAC address the resulting IPv6 address would be: 5fbc:1000::0a:0b:0c:0d:0e:0f > another question. > Can someone explain me the following adresses : > - my ipv6 adress > See above > - a link ipv6 adress (entry of tunnel) > - a link ipv6 adress (end of tunnel) > - the tunnel endpoint ipv6 adress > A tunnel interface is a pseudo interface that has as parameters a pair of IPv6 addresses (corresponding to the tunnel entry point node and tunnel exit point node). The pseudo-interface is layered on the top of a real interface. A tunnel's entry point IPv6 address is one of the underlying interface's IPv6 addresses. This is the address that will be filled as IPv6 source address in the tunnel header. A tunnel's exit point address is remote node IPv6 address. It is filled in as destination IPv6 address in the tunnel IPv6 address. On transmit the tunnel packets are send to the tunnel exit point node address, and thus forwarded to the next hop router that would route the packet to its destination. I hope this helps, Alex > Peace > > Hamdi > --------------------------------------------- Alex Conta aconta@lucent.com Tel: 508/287-9000/ext 2842 Fax: 508/287-2810 Lucent Technologies, Inc. 300 Baker Ave., Concord, MA 01742 From Alain.Durand@imag.fr Wed Jun 11 15:04:35 1997 From: Alain.Durand@imag.fr (Alain Durand) Date: Wed, 11 Jun 1997 16:04:35 +0200 Subject: new addressing plan In-Reply-To: Pedro Marques "Re: new addressing plan" (Jun 10, 3:22pm) References: <"199706101634.JAA14680"@trix.cisco.com> <199706102159.QAA15225@gungnir.fnal.gov> <199706102222.PAA28016@pedrom-ultra.cisco.com> Message-ID: <970611160435.ZM23587@rama.imag.fr> On Jun 10, 3:22pm, Pedro Marques wrote: > Subject: Re: new addressing plan > But anyway from the sample on the mailing list it seams people prefer a > registry delegation based scheme to an automatic one... Pedro, automatic allocation was fine when we started the 6-bone. At the time, there was only a small registry, the RIPE databased. This automatic allocation allowed everybody to start quickly a small IPv6 tesbed and then try to connect to the 6-bone. Now, it's a different game. The 6-bone is there, we have a registry. One of the goal of the 6-bone (as a group) is to simulate IPv6 providers to test the new addressing plan. I'd rather like to see the registry allocating new prefixes to core nodes and those sites re-allocating longer prefixes to transit & leaf nodes We could then test site renumbering, provider renumbering, aggregation, etc... To do that, there is no need to use AS nubers nor IPv4 numbers to build v6 addreses. I think we can rely on our registry and some sub-registries to do the work. - Alain. From Thomas.Harsch@unisys.com Wed Jun 11 16:20:47 1997 From: Thomas.Harsch@unisys.com (Harsch, Tom) Date: Wed, 11 Jun 1997 09:20:47 -0600 Subject: ipv6 Prefix and local adress Message-ID: Alex, > >No the prefix of your link local address is a predefined hardcoded >value: > > FFFE:8000::0/18 ^^^^^^^^^^^^^^^ Is this correct? RFC-1844 and indicate "FE80::0/10". -- Tom Harsch Unisys, Salt Lake City, UT mailto:Thomas.Harsch@UNISYS.com mailto:tcharsch@acm.org > > > From AConta@lucent.com Wed Jun 11 18:19:27 1997 From: AConta@lucent.com (Conta, Alex) Date: Wed, 11 Jun 1997 13:19:27 -0400 Subject: ipv6 Prefix and local adress Message-ID: Dan Harrington pointed out to me (thanks Dan) a mistake (sorry) in the information I sent out last night on this list. The correct information is: The prefix of your link local address is a predefined hardcoded > value: > > FE80::/64 > Thanks, Alex --------------------------------------------- Alex Conta aconta@lucent.com Tel: 508/287-9000/ext 2842 Fax: 508/287-2810 Lucent Technologies, Inc. 300 Baker Ave., Concord, MA 01742 > ---------- > From: Conta, Alex > Sent: Tuesday, June 10, 1997 7:24 PM > To: 6bone@ISI.EDU > Subject: RE: ipv6 Prefix and local adress > > > > Hello, > > someone at isi gave me the prefix : 5fbc:1000::0/32 > > should this be the prefix of my local ipv6 adress too ? > > > No the prefix of your link local address is a predefined hardcoded > value: > > FFFE:8000::0/18 > From Thomas.Harsch@unisys.com Wed Jun 11 20:44:26 1997 From: Thomas.Harsch@unisys.com (Harsch, Tom) Date: Wed, 11 Jun 1997 13:44:26 -0600 Subject: ipv6 Prefix and local adress Message-ID: Alex, > >Alex, >> >>No the prefix of your link local address is a predefined hardcoded >>value: >> >> FFFE:8000::0/18 > ^^^^^^^^^^^^^^^ >Is this correct? RFC-1844 and Oops! I meant RFC-1884. >indicate "FE80::0/10". -- Tom Harsch Unisys, Salt Lake City, UT mailto:Thomas.Harsch@UNISYS.com mailto:tcharsch@acm.org > From davek@research.att.com Thu Jun 12 02:09:48 1997 From: davek@research.att.com (Dave Kormann) Date: Wed, 11 Jun 1997 21:09:48 -0400 (EDT) Subject: RIPE ftp group password? Message-ID: <199706120109.VAA26657@radish.research.att.com> so, i've consed up a RIPE entry for my site, and i'd like to drop it in the registry; the web pages mention a group password, and that the password should be obtained from the mailing list. i guess this is a request for that, then; anyone who could give me pointers would help. thanks! oh, here, fwiw, is the RIPE entry. is this sane? i didn't include a location-string 'cause i haven't got a GPS handy :) site: AT&T Labs - Research Weblab location: AT&T Labs - Research, Florham Park, NJ prefix: 5f00:100:cf8c:a800:a8::/80 ping: 5f00:100:cf8c:a800:a8:800:2080:f296 tunnel 207.140.168.50 132.177.118.22 UNH (RIPng) contact: weblab@research.att.com status: operational since 04-Jun-1997 remark: We're still pretty flaky at the moment, but willing remark: to supply tunnels to people who want to risk it. remark: Planning to run an HTTPD at some time in the near future. remark: End system: Sun Ultra Enterprise 150, Solaris 2.5.1 changed: davek@research.att.com 970611 source: RIPE Dave Kormann AT&T Labs Dash Research From davidk@isi.edu Thu Jun 12 18:36:17 1997 From: davidk@isi.edu (davidk@isi.edu) Date: Thu, 12 Jun 1997 10:36:17 -0700 (PDT) Subject: RIPE ftp group password? In-Reply-To: <199706120109.VAA26657@radish.research.att.com> from "Dave Kormann" at Jun 11, 97 09:09:48 pm Message-ID: <9706121736.AA24404@brind.isi.edu> Dave, Dave Kormann writes: > > so, i've consed up a RIPE entry for my site, and i'd like to drop it > in the registry; the web pages mention a group password, and that the > password should be obtained from the mailing list. i guess this is a > request for that, then; anyone who could give me pointers would help. > thanks! We recently changed to a new registry format as described on the bottom of the 6bone web page. The web page is not updated yet due to a vacation of the maintainer. Please check out the 'possible new registry links' at the bottom of the home page. Let me know in a private mail if you want me to convert your entry as shown below to the new format to save you some work. Don't hesitate to send me a mail if you have more questions, > oh, here, fwiw, is the RIPE entry. is this sane? i didn't include a > location-string 'cause i haven't got a GPS handy :) This is fine. It's optional although it would be nice if you knew the location or could approximate it using a map. David K. --- > site: AT&T Labs - Research Weblab > location: AT&T Labs - Research, Florham Park, NJ > prefix: 5f00:100:cf8c:a800:a8::/80 > ping: 5f00:100:cf8c:a800:a8:800:2080:f296 > tunnel 207.140.168.50 132.177.118.22 UNH (RIPng) > contact: weblab@research.att.com > status: operational since 04-Jun-1997 > remark: We're still pretty flaky at the moment, but willing > remark: to supply tunnels to people who want to risk it. > remark: Planning to run an HTTPD at some time in the near future. > remark: End system: Sun Ultra Enterprise 150, Solaris 2.5.1 > changed: davek@research.att.com 970611 > source: RIPE > > Dave Kormann > AT&T Labs Dash Research > David K. --- From bmanning@ISI.EDU Fri Jun 13 22:12:43 1997 From: bmanning@ISI.EDU (bmanning@ISI.EDU) Date: Fri, 13 Jun 1997 14:12:43 -0700 (PDT) Subject: ipv6 Prefix and local adress In-Reply-To: <339DA37A.BBF@ati.tn> from "Hamdi TOUNSI" at Jun 10, 97 05:56:58 pm Message-ID: <199706132112.AA24766@zed.isi.edu> > > Hello, > someone at isi gave me the prefix : 5fbc:1000::0/32 > should this be the prefix of my local ipv6 adress too ? > another question. > Can someone explain me the following adresses : > - my ipv6 adress > - a link ipv6 adress (entry of tunnel) > - a link ipv6 adress (end of tunnel) > - the tunnel endpoint ipv6 adress > > Peace > > Hamdi Er, no. That is my prefix. Your prefix should be created using the methods specified in the rfcs. The tunnel endpoints are calculated based on the prefix in use. -- --bill From roque@cisco.com Sat Jun 14 03:39:17 1997 From: roque@cisco.com (Pedro Marques) Date: Fri, 13 Jun 1997 19:39:17 -0700 (PDT) Subject: ipv6 Prefix and local adress In-Reply-To: <199706132112.AA24766@zed.isi.edu> References: <339DA37A.BBF@ati.tn> <199706132112.AA24766@zed.isi.edu> Message-ID: <199706140239.TAA29792@pedrom-ultra.cisco.com> >>>>> "Bill" == bmanning writes: >> Hello, someone at isi gave me the prefix : 5fbc:1000::0/32 >> should this be the prefix of my local ipv6 adress too ? >> another question. Can someone explain me the following >> adresses : - my ipv6 adress - a link ipv6 adress (entry of >> tunnel) - a link ipv6 adress (end of tunnel) - the tunnel >> endpoint ipv6 adress >> >> Peace >> >> Hamdi Bill> Er, no. That is my prefix. Your prefix should be created Bill> using the methods specified in the rfcs. The tunnel Bill> endpoints are calculated based on the prefix in use. Bill, you could delegate him a /64 out of your /32... If he is going to connect to you there is no reason why he should use it own prefix, i believe. Pedro. From bmanning@ISI.EDU Mon Jun 16 15:26:50 1997 From: bmanning@ISI.EDU (Bill Manning) Date: Mon, 16 Jun 1997 07:26:50 -0700 (PDT) Subject: ipv6 Prefix and local adress In-Reply-To: <199706140239.TAA29792@pedrom-ultra.cisco.com> from "Pedro Marques" at Jun 13, 97 07:39:17 pm Message-ID: <199706161426.AA19495@zephyr.isi.edu> > > >>>>> "Bill" == bmanning writes: > > >> Hello, someone at isi gave me the prefix : 5fbc:1000::0/32 > >> should this be the prefix of my local ipv6 adress too ? > >> another question. Can someone explain me the following > >> adresses : - my ipv6 adress - a link ipv6 adress (entry of > >> tunnel) - a link ipv6 adress (end of tunnel) - the tunnel > >> endpoint ipv6 adress > >> > >> Peace > >> > >> Hamdi > > Bill> Er, no. That is my prefix. Your prefix should be created > Bill> using the methods specified in the rfcs. The tunnel > Bill> endpoints are calculated based on the prefix in use. > > Bill, you could delegate him a /64 out of your /32... > If he is going to connect to you there is no reason why he should use > it own prefix, i believe. > > Pedro. True, although the current model for address allocation is not aggregatable outside the AS. And since he is not within my AS and we are not yet ready to transition to the proposed alternate format, I am not willing to carve out a chunk of space for him from my block(s). This is more for the upstream reporting than anything else. And there is every likleyhood that he will connect elsewhere (punching a hole in the block). -- --bill From RLFink@lbl.gov Thu Jun 26 02:11:33 1997 From: RLFink@lbl.gov (Bob Fink) Date: Wed, 25 Jun 1997 18:11:33 -0700 Subject: 6bone diagrams and ISI RR pointer changes Message-ID: Hi! I'm back, fully rested, from three weeks of vacation in glorious Italy! Many thanks to David Kessens for converting over to the new RIPE-style RR db. I have changed the RR data hot buttons for the two 6bone diagams to use the new database. Please let me know if there is a problem with anything. Thanks, Bob