Re: Stupid NAT tricks and how to stop them.
Why would a service provider give up skimming the cream with that (nearly free) extra cash that weirdos like us hand them for real IPv4 addresses? Eliot ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
Hi Andrew, And people wonder why NATs proliferate... much of the world has no option but to live with them. This is a direct result of policy discouraging IPv4 address allocation. sorry for asking, but what policy are you referring to? RIR policy? Can you point out any RIRs policy that prevents from getting one public IPv4 address per machine connected to the Internet? What do you think that needs to be changed in the v4 allocation policy? Or are you talking about business model of the ISPs? (which doesn't seem to me to be related with policies, but just business...) Thanks, marcelo Andrew ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
On 30-mrt-2006, at 10:29, marcelo bagnulo braun wrote: And people wonder why NATs proliferate... much of the world has no option but to live with them. This is a direct result of policy discouraging IPv4 address allocation. sorry for asking, but what policy are you referring to? RIR policy? Can you point out any RIRs policy that prevents from getting one public IPv4 address per machine connected to the Internet? On a somewhat (un)related note: it's not easy for ISPs to give out two or three IP addresses to customers because there is no good mechanism to do so. One address works very well with PPP or DHCP, but a specific number other than one doesn't, so the next step is something like a /29. On an even more (un)related note: it's not possible to give IPv6 addresses to customers over PPP, and it's very inconvenient to make a /64 - /48 be routed towards a customer router that dynamically connects to an ISP network. (I.e., my cell phone is a router and it dials up, it gets an address through stateless autoconfig but then my laptop, PDA etc that use the cell phone as their router aren't automatically reachable.) Some more work on provisioning mechanisms wouldn't be a bad thing. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: Stupid NAT tricks and how to stop them.)
Dear Illjitsch, full agreement with everything you say. In addition please consider that engineers want to square several informations within a single space. So to the lose of numbering space they add rigidity wich operationnally kills still more space. Actually there is a 32 Hexa long address space. Question 1 is how many Hexa to protocol, how many to routing, how many to addressing and how many to subaddressing. Question 2 are the solutions to save space depending on the use within that ABNF without removing operational simplicity. It seems absurd to give my mobile as many local addresses than to a large corporation network. Question 3 is interoperability of the various addressing spaces. Anyone having to build a numbering plan for any kind of nomenclature has the same problem and faces the same erroneous propositions. I wander if someone wrote a book about that kind of experience the world could take advantage from. The problem is that this incertainity blocks network application needind a stable relative numbering plans. jfc At 06:26 30/03/2006, Anthony G. Atkielski wrote: Iljitsch van Beijnum writes: > So how big would you like addresses to be, then? ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: Stupid NAT tricks and how to stop them.)
On 30-mrt-2006, at 6:26, Anthony G. Atkielski wrote: We currently have 1/8th of the IPv6 address space set aside for global unicast purposes ... Do you know how many addresses that is? One eighth of 128 bits is a 125-bit address space, or 42,535,295,865,117,307,932,921,825,928,971,026,432 addresses. That's enough to assign 735 IP addresses to every cubic centimetre in the currently observable universe (yes, I calculated it). Am I the only person who sees the absurdity of wasting addresses this way? When I first learned about IPv6 I felt strongly that 128 bits was too much, especially since all those bits have to be carried in every IP packet twice, once as a source address and once as a destination address. However, since that time I've learned to appreciate stateless autoconfiguration and the potential usefulness of having the lower 64 bits of the IPv6 address as a place to carry some limited security information (see SEND and shim6 HBA). ... with the idea that ISPs give their customers /48 blocks. Thank you for illustrating the classic engineer's mistake. Stop thinking in terms of _bits_, and think in terms of the _actual number of addresses_ available. Of better still, start thinking in terms of the _number of addresses you throw away_ each time you set aside entire bit spans in the address for any predetermined purpose. The trouble is that you need to build in space for growth. Unfortunately, at the time IPv6 was created variable length addresses weren't considered viable. (In theory CLNP has variable length addresses, in practice that doesn't really work out.) And for some strange reason, apparently only powers of two were considered as address lengths. So the choice was either 64 bits, which is a lot, but it doesn't allow for any innovation over the 32 bits we have in IPv4, or 128 which does cost 16 extra bytes in every packet, but gives us stateless autoconfig and SEND. Now you can argue that 64 or 48 bits and continuing current IPv4 practices would have been better, but given the choice for 128 bits, the current way of using the address space makes sense for the most part. The only thing I'm not too happy about is the current one address / one subnet / /48 trichotomy. Ignoring the single address for a moment, the choice between one subnet and 65536 isn't a great one, as many things require a number of subnets that's greater than one, but not by much. For instance, the cell phone as a router example I talked about earlier. A /64 and a single address, or two /64s which would be a /63 would be more useful there. The idea that we'd use up too much address space by giving out /48s doesn't seem like a real problem to me, but on the other hand most people don't need a /48 so some choice thats < /48 and > 64 makes sense. But a /56 as suggested by some people is suboptimal: people with growing networks will at some point need more than 256 subnets but at that point they already have very many subnets so renumbering then is painful. Making the choice between /60 and /48 makes much more sense: just give everyone a /60 rather than a /64 just in case they need a handful of subnets, which is adequate for 99% of all internet users. People who need more than a handful of subnets can get a /48 and won't have to renumber at an inconvenient point in the growth curve. Yes, I know this will use up sixteen times the number of addresses for people who really only need a single subnet, but it saves a factor 4096 for the people who need 2 - 16 subnets and would have gotten a /48 or a factor 16 for the people who would have gotten a / 56. The extra wasted address space from giving people who really only need one subnet a /60 is made up for by the people who would have gotten a /48 but can now get by with a /60 if the ratio of one-subnet to <17-subnet users is 1 : 4000 or less. (Making the choice /64 vs / 56 saves even more as long as the ratio is lower than 94 : 6 but doesn't have the desireable near-onesizefitsall quality.) If you want exponential capacity from an address space, you have to assign the addresses consecutively and serially out of that address space. You cannot encode information in the address. You cannot divided the address in a linear way based on the bits it contains and still claim to have the benefits of the exponential number of addresses for which it supposedly provides. The thing that is good about IPv6 is that once you get yourself a / 64, you can subdivide it yourself and still have four billion times the IPv4 address space. (But you'd be giving up the autoconfiguration advantages.) Also, when the time comes to create the next version of IP, we won't have to worry about all of this to a noticeable degree because IPvA or IPvF or whatever can have a different addressing structure that can still be expressed as an IPv6-like 128 bit number for backward compatibility with
Re: Stupid NAT tricks and how to stop them.
On 28 mar 2006, at 18.00, Hallam-Baker, Phillip wrote: From: Kurt Erik Lindqvist [mailto:[EMAIL PROTECTED] NAT is a dead end. If the Internet does not develop a way to obsolete NAT, the Internet will die. It will gradually be replaced by networks that are more-or-less IP based but which only run a small number of applications, poorly, and expensively. ...or you will see an overlay network build on top of NAT+IPv4 that abstracts the shortcomings away - aka what the peer to peer networks are doing. End-to-end addressing... Precisely. Just what is this fetish about keeping the IP address the same as the packet travels? I will have to get better at making irony clearerI most certainly hope we are not heading down the route I suggest above. I am _afraid_ we are though. - kurtis - ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: Stupid NAT tricks and how to stop them.)
On Thu, Mar 30, 2006 at 01:36:18PM +0200, Iljitsch van Beijnum wrote: > > The thing that is good about IPv6 is that once you get yourself a / > 64, you can subdivide it yourself and still have four billion times > the IPv4 address space. (But you'd be giving up the autoconfiguration > advantages.) I noticed that by deafult MS Vista doesn't use autoconf as per 2462, rather it uses a 3041-like random address. See: http://www.microsoft.com/technet/itsolutions/network/evaluate/new_network.mspx "Random Interface IDs for IPv6 Addresses To prevent address scans of IPv6 addresses based on the known company IDs of network adapter manufacturers, Windows Server "Longhorn" and Windows Vista by default generate random interface IDs for non-temporary autoconfigured IPv6 addresses, including public and link-local addresses." That reads to me like no 2462 by default. Maybe I'm misinterpreting. One could envisage an option where that randomness is applied to 48 host bits not 64. If you really really wanted to do that. -- Tim/::1 ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
RE: 128 bits should be enough for everyone, was:
Anthony, The problem with allocating numbers sequentially is the impact on routers and routing protocols. I have heard that the Japanese issue house numbers chronologically. When you find the right block, you have to hunt for the right number. What you are suggesting is similar. You would have as many routing table entries as hosts in the world. The router would not be affordable. The traffic for routing entries would swamp the net. The processing of these routing advertisements would be impossible. It doesn't scale! The function of an address is to enable a router to find it. That is why we try to use hierarchical addressing even at the cost of numbering space. IMO one problem of the Internet is that it isn't hierarchical enough. Consider the phone system: country codes, area codes ... This makes the job of building a switch much easier. I think we should have divided the world into 250 countries. Each country into 250 "provinces". Yes, it would waste address space but it would make routing much easier and more deterministic. It would simplify BGP drastically. Current routing algorithms aren't that efffective. Paths in the net are fairly inefficient. This increases latency, apparent traffic on the net, and is a real waste of money. Yes this would mean a mobile node needs to get new addresses as it moves. So what. We already have DHCP. Cell phones do a handoff already. Steve Silverman > -Original Message- > From: Anthony G. Atkielski [mailto:[EMAIL PROTECTED] > Sent: Wednesday, March 29, 2006 11:27 PM > To: ietf@ietf.org > Subject: Re: 128 bits should be enough for everyone,was: > IPv6 vs. Stupid > NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and > how to stop > them.) > > > Iljitsch van Beijnum writes: > > > So how big would you like addresses to be, then? > > It's not how big they are, it's how they are allocated. > And they are > allocated very poorly, even recklessly, which is why they run out so > quickly. It's true that engineers always underestimate required > capacity, but 128-bit addresses would be enough for anything ... IF > they were fully allocated. But I know they won't be, and so the > address space will be exhausted soon enough. > > > We currently have 1/8th of the IPv6 address space set aside for > > global unicast purposes ... > > Do you know how many addresses that is? One eighth of 128 bits is a > 125-bit address space, or > > 42,535,295,865,117,307,932,921,825,928,971,026,432 > > addresses. That's enough to assign 735 IP addresses to every cubic > centimetre in the currently observable universe (yes, I calculated > it). Am I the only person who sees the absurdity of wasting > addresses > this way? > > It doesn't matter how many bits you put in an address, if you assign > them this carelessly. > > > ... with the idea that ISPs give their customers /48 blocks. > > Thank you for illustrating the classic engineer's mistake. Stop > thinking in terms of _bits_, and think in terms of the > _actual number > of addresses_ available. Of better still, start thinking > in terms of > the _number of addresses you throw away_ each time you set aside > entire bit spans in the address for any predetermined purpose. > > Remember, trying to encode information in the address (which is what > you are doing when you reserve bit spans) results in > exponential (read > incomprehensibly huge) reductions in the number of available > addresses. It's trivially easy to exhaust the entire address space > this way. > > If you want exponential capacity from an address space, you have to > assign the addresses consecutively and serially out of that address > space. You cannot encode information in the address. You cannot > divided the address in a linear way based on the bits it > contains and > still claim to have the benefits of the exponential number of > addresses for which it supposedly provides. > > Why is this so difficult for people to understand? > > > That gives us 45 bits worth of address space to use up. > > You're doing it again. It's not 45 bits; it's a factor of > 35,184,372,088,832. > > But rest assured, they'll be gone in the blink of an eye if the > address space continues to be mismanaged in this way. > > > It's generally accepted that an HD ratio of 80% should be > reachable > > without trouble, which means we get to waste 20% of those > bits in > > aggregation hierarchies. > > No. It's not 20% of the bits, it's 99.9756% of your address > space that > you are wasting. > > Do engineers really study math? > > > This gives us 36 bits = 68 billion /48s. That's several per person > > inhabiting the earth, and each of those / 48s provides > 65536 subnets > > that have room to address every MAC address ever assigned without > > breaking a sweat. > > What happens when MAC addresses go away? How are you providing for > the future when you allocate address space based on the > past? Why not > just leave the address space alone, and allocate only the minimum > slice
OT: Japanese addresses (was: 128 bits should be enough for everyone)
Steve Silverman [mailto:[EMAIL PROTECTED] writes: > I have heard that the Japanese issue house numbers chronologically. > When you find the right block, you have to hunt for the right number. It's even worse. An address written "Latin" characters is typically in the form Bldg#-Block#-Area# DistrictName, WardName, CityName PostCode#, or sometimes Bldg#-Block#, DistrictName Area#-chome, WardName, CityName PostCode#. Bldg# is chronological. Area# does not follow any geographical pattern. I don't know about Block# or PostCode#. Note the lack of a street name in the address; that's because most non-major streets don't HAVE names! I've been running into this problem, trying to find on a map the building in Tokyo where I'll be working in mid-April (10-19). Anybody near either Shiba or Akasaka want to get together for a biiru (or kohi or whatever) then? -Dave ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
--On Thursday, March 30, 2006 19:30 +1200 Andrew McGregor <[EMAIL PROTECTED]> wrote: Your ISP charges you 9 times as much for IPv4 addresses as they do for bandwidth? I'd recommend switching ISPs. All the ones I've seen charge a small premium for additional IP space, but it's never more than about a 50% premium. Not if you don't live in the US. There are no options here that are at all cheap. Usually you get a flat "we don't do that". And they don't do v6 either. If it makes you feel better (it probably won't), in much of the US, the story from the ISPs goes like this: * We don't do that on our residential service, if you want _any_ IPv4 addresses assigned to you, you need to buy the commercial service. * The commercial service costs around ten times as much as the residential one for similar bandwidth _less_ service (often no free email, free web hosting, "user protection" software tools, etc.) * If you want more than one address on the commercial service, you will pay some small incremental charge for it. But the real incremental charge starts at address number 1 and is tied up with the "type of service" shift. However, we need to keep something else in mind, which Iljitsch's note hints at. If I'm an ISP trying to sell a low-end service to low-end customers at a low (but still profitable) price, I need to cut customer support costs to the absolute minimum. If someone calls up for help with a configuration problem, that may be six month's of profits from that customer eaten up in the cost of answering the call. To that sort of ISP, NATs, and ISP-supplied routers that support NATs, have a _huge_ advantage, which is that all supported customer LANs are identical -- same design, same exact internal addresses, etc.That is very important from a support standpoint -- length of calls, skill levels required, ability to construct clear FAQs and avoid calls entirely, and so on. For the community, there are elements of "you get what you pay for" in this. And, for the ISPs, unless we figure out ways to provide the same level of support convenience with public addresses, we will certainly see NATs with IPv6 as well as IPv4. john ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Last Call: 'TLS User Mapping Extension' to Proposed Standard
http://www.ietf.org/internet-drafts/draft-santesson-tls-ume-04.txt says: This document does not specify how the server stores the user_principal_name, or how exactly it might be used to locate a certificate. For instance, it might be appropriate to do a case- insensitive lookup. It is RECOMMENDED that the server processes the user_principal_name with a stringprep profile [N7] appropriate for the identity in question, such as Nameprep [N8] for the portion domain portion of UPN, SASLprep [N9] for the user portion of the UPN and stringprep appendix B.3 [N7] as mapping table for case folding. Given that the first and second sentence make it clear that the use of StringPrep is not required, I suggest using MAY instead of RECOMMENDED in the third sentence. RECOMMENDED is the same as SHOULD according to RFC 2119, and is a fairly strong recommendation. Its use seem misplaced here. It may be better to avoid RFC 2119 language completely here, because the entire paragraph is merely an example of what you can do. Thanks, Simon The IESG <[EMAIL PROTECTED]> writes: > The IESG has received a request from an individual submitter to consider the > following documents: > > - 'TLS User Mapping Extension' > as a Proposed Standard > - 'TLS Handshake Message for Supplemental Data' > as a Proposed Standard > > The previous Last Call on draft-santesson-tls-ume-03.txt has finished. > However, to resolve some comments that were received during the > previous Last Call, the document has been updated and > draft-santesson-tls-supp-00.txt was written. Due to the significant > changes in one area of the document, the IESG is making a second > call for comments. This comment period is shorter since the majority > of the document is unchanged. > > The IESG plans to make a decision in the next few weeks, and solicits > final comments on this action. Please send any comments to the > iesg@ietf.org or ietf@ietf.org mailing lists by 2006-04-11. > > The file can be obtained via > http://www.ietf.org/internet-drafts/draft-santesson-tls-ume-04.txt > http://www.ietf.org/internet-drafts/draft-santesson-tls-supp-00.txt ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
However, we need to keep something else in mind, which Iljitsch's note hints at. If I'm an ISP trying to sell a low-end service to low-end customers at a low (but still profitable) price, I need to cut customer support costs to the absolute minimum. If someone calls up for help with a configuration problem, that may be six month's of profits from that customer eaten up in the cost of answering the call. I find myself wondering, don't they get support calls from customers having to deal with the problems caused by the NATs? For the community, there are elements of "you get what you pay for" in this. And, for the ISPs, unless we figure out ways to provide the same level of support convenience with public addresses, we will certainly see NATs with IPv6 as well as IPv4. either that, or IPv6 will be seen as something that is "business use only". Keith ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
> If someone calls up for help with a > configuration problem, that may be six month's of > profits from that customer eaten up in the cost of answering the call. That is because the current Internet pricing has been screwed-up from the start. LD settlements between telcos are fully applicable to ISPs but have never been instituted. Internet has been subsidised for years by the local access but now as wireline declines everybody starts feeling the pain. Usage based billing and inter-ISP settlements start showing up lately and they fit well for the Internet. Otherwise transit providers as well as heavy users rip all the benefits. Peter Sherbin --- John C Klensin <[EMAIL PROTECTED]> wrote: > > > --On Thursday, March 30, 2006 19:30 +1200 Andrew > McGregor > <[EMAIL PROTECTED]> wrote: > > >> Your ISP charges you 9 times as much for IPv4 > addresses as > >> they do for bandwidth? I'd recommend switching > ISPs. All > >> the ones I've seen charge a > >> small premium for additional IP space, but it's > never more > >> than about a 50% premium. > > > > Not if you don't live in the US. There are no > options here > > that are at all cheap. Usually you get a flat > "we don't do > > that". And they don't do v6 either. > > If it makes you feel better (it probably won't), in > much of the > US, the story from the ISPs goes like this: > > * We don't do that on our residential service, if > you > want _any_ IPv4 addresses assigned to you, you need > to > buy the commercial service. > > * The commercial service costs around ten times as > much > as the residential one for similar bandwidth _less_ > service (often no free email, free web hosting, > "user > protection" software tools, etc.) > > * If you want more than one address on the > commercial > service, you will pay some small incremental charge > for > it. But the real incremental charge starts at > address > number 1 and is tied up with the "type of service" > shift. > > However, we need to keep something else in mind, > which > Iljitsch's note hints at. If I'm an ISP trying to > sell a > low-end service to low-end customers at a low (but > still > profitable) price, I need to cut customer support > costs to the > absolute minimum. If someone calls up for help with > a > configuration problem, that may be six month's of > profits from > that customer eaten up in the cost of answering the > call. To > that sort of ISP, NATs, and ISP-supplied routers > that support > NATs, have a _huge_ advantage, which is that all > supported > customer LANs are identical -- same design, same > exact internal > addresses, etc.That is very important from a > support > standpoint -- length of calls, skill levels > required, ability to > construct clear FAQs and avoid calls entirely, and > so on. > > For the community, there are elements of "you get > what you pay > for" in this. And, for the ISPs, unless we figure > out ways to > provide the same level of support convenience with > public > addresses, we will certainly see NATs with IPv6 as > well as IPv4. > > john > > > > ___ > Ietf mailing list > Ietf@ietf.org > https://www1.ietf.org/mailman/listinfo/ietf > __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
Andrew McGregor writes: > Not if you don't live in the US. There are no options here that are > at all cheap. Usually you get a flat "we don't do that". And they > don't do v6 either. True. > And people wonder why NATs proliferate... much of the world has no > option but to live with them. This is a direct result of policy > discouraging IPv4 address allocation. And a rather skewed allocation of address space, more evidence of engineers not knowing what they are doing. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
On Thu, Mar 30, 2006 at 11:44:37AM -0500, Keith Moore wrote: > >However, we need to keep something else in mind, which Iljitsch's note > >hints at. If I'm an ISP trying to sell a low-end service to low-end > >customers at a low (but still profitable) price, I need to cut customer > >support costs to the absolute minimum. If someone calls up for help > >with a configuration problem, that may be six month's of profits from > >that customer eaten up in the cost of answering the call. > > I find myself wondering, don't they get support calls from customers > having to deal with the problems caused by the NATs? No, because most customers browse the web, read e-mail, use skype/VOIP services, and all of those work under NAT. A number of VPN packages break, (a) that's not that common compared the huge number of residential customers that aren't doing VPN's for work-at-home setups, andin any case, those complaints usually gets sent to the company help desk, and a number of VPN's have solutions that work with NAT's anyway. The problems caused by NAT's are the sort of things that don't normally show up at ISP help desks; the new applications that are not written, the architectures that are torqued to deal with NAT's, etc. - Ted ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: Stupid NAT tricks and how to stop them.)
Iljitsch van Beijnum writes: > When I first learned about IPv6 I felt strongly that 128 bits was too > much, especially since all those bits have to be carried in every IP > packet twice, once as a source address and once as a destination > address. When I first learned about IPv6 I started worrying that it wouldn't last for very long before being exhausted in turn. And I worried a lot more when I saw the mistakes of the past being repeated--the very mistakes that wasted so many IPv4 addresses. > However, since that time I've learned to appreciate > stateless autoconfiguration and the potential usefulness of having > the lower 64 bits of the IPv6 address as a place to carry some > limited security information (see SEND and shim6 HBA). Once it's carrying information, it's no longer just an address, so counting it as pure address space is dangerous. > The trouble is that you need to build in space for growth. You build in space for growth by not trying to allocate address spaces in advance. For example, you allocate 0...0+33 bits for new addresses, and now you've doubled the IPv4 address space (and bought yourself years of additional time), while using up only an infinitesimal portion of the IPv6 space. But when you start chopping the address into sections, you throw almost all the address space away ... and when that happens, you are going to exhaust that space in no time, no matter how many bits it has. Building in space means not allocating it--not even _planning_ to allocate it. Nobody has any idea what the Internet might be like a hundred years from now, so why are so many people hellbent on "planning" for something they can't even imagine? > Unfortunately, at the time IPv6 was created variable length addresses > weren't considered viable. Variable-length addresses are the only permanent solution, unless IP addresses are assigned serially (meaning that all routing information has to be removed). Variable-length addresses work very well for the telephone system, and they'd work just as well for the Internet, if only someone had taken the time to work it out. > The only thing I'm not too happy about is the current one address / > one subnet / /48 trichotomy. Ignoring the single address for a > moment, the choice between one subnet and 65536 isn't a great one, as > many things require a number of subnets that's greater than one, but > not by much. It's a good example of waste that results from short-sightedness. It happened in IPv4, too. > The thing that is good about IPv6 is that once you get yourself a / > 64, you can subdivide it yourself and still have four billion times > the IPv4 address space. It sounds like NAT. > I'm not a huge fan of the HD ratio either, because it substitues a > rule of thumb for actual knowledge. But the point is that EVEN if you > waste 99.9756% in this way we STILL have enough addresses to give > every person living on the planet when the population hits its peak > several /48s which are wasteful in their own right. Famous last words. I've seen virtual memory systems run out of virtual address space, even when that space contained (in theory) more bytes than anyone could ever possibly build into any real-world system. The reason? Careless allocation of the addresses. No matter how many bits you have, you can blow through them in linear time if you allocate them based on bit fields, and it seems that virtually no engineers can resist the urge to do exactly that. > So while I wouldn't want to take away your right to begrudge the way > all of this is done in IPv6, I must object to your conclusion that > we'll run out of IPv6 soon, for any reasonable value of "soon". Well, time will tell, won't it? > I hope good engineers don't think that ... Any engineer setting aside bit spans in an address for future use is thinking exactly that, and he'll be wrong. > Engineers should build stuff that still works reasonably well even if > they get their predictions wrong. Engineers don't like to think that they've left anything out or that they are less than omniscient in assessing what must be done, so many of them are allergic to anything that is simply "reserved for future use." I had the same trouble when I first started in computers, but I grew out of it. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
At 17:44 30/03/2006, John C Klensin wrote: If I'm an ISP trying to sell a low-end service to low-end customers at a low (but still profitable) price, I need to cut customer support costs to the absolute minimum. Interesting to see the impact of a longer datacom societal culture. This may give an indication on the future US trend? In France you pay the phone support by the minute. A telephone charge system inherited from Minitel, and "Audiotel". There are different rates which must be advertised. The on-line wait of the support was charged. Since Jan 1st, you pay the time you get actual support on line. Rate for free telephone nationwide + ADSL is 24.5 euro or less. There start being free telephone international. Rates use to be 29.95 euro when including either TV or free support. Dedicated IPv4 are possible. Specialised pro ISP offer IPv6/IPv4 ADSL for around twice that price plus free email support and real hosting and emails. You can chose not to pay a Telco anymore, just the ISP. After six months with a Telco which installs the line. This is transition while ISP have not organised the line installation or WI-Max access. The interest is certainly to provide bundled IPv6 numbers for free. This was the underlying intent in a questionnaire of the Regulation Authority last year. Increasingly we use coreboxes (smart home routers, NAT+ by ISPs). The trend is probably towards an interbox network as they become more and more intelligent and powerful. Most of the mobile subscription offer one to three free numbers. So you can plug your home number on the net, make you home number free from you mobile and phone to fixed phones for free. I suppose we will soon see a box permitting to establish conference calls whcih will permit to call mobiles for far less than an SMS. jfc ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was:
Steve Silverman writes: > The problem with allocating numbers sequentially is the impact on > routers and routing protocols. The problem with not doing so is that a 128-bit address doesn't provide anything even remotely close to 2^128 addresses. You have to choose what you want. > I have heard that the Japanese issue house numbers chronologically. > When you find the right block, you have to hunt > for the right number. What you are suggesting is similar. You would > have as many routing table entries as hosts in the world. The router > would not be affordable. The traffic for routing entries would swamp > the net. The processing of these > routing advertisements would be impossible. It doesn't scale! Variable address length scales, and it never runs out of addresses, but nobody wants to do that, even though telephones have been doing it for ages. > The function of an address is to enable a router to find it. That is > why we try to use hierarchical addressing even at the cost of numbering > space. In that case, assign addresses to points in space, instead of devices. An office occupying a given plot of land will have an IP address space that is solely a function of the space it occupies. Routing would be the essence of simplicity and blazingly fast. > IMO one problem of the Internet is that it isn't hierarchical enough. > Consider the phone system: country codes, area codes ... This makes > the job of building a switch much easier. I think we should have > divided the world into 250 countries. Each country into 250 > "provinces". Yes, it would waste address space but it would make > routing much easier and more deterministic. With a variable address length that can extend infinitely at either end, the address space would never be exhausted. That's how telephones work. > Yes this would mean a mobile node needs to get new addresses as it > moves. So what. We already have DHCP. Cell phones do a handoff > already. I agree. We also have DNS. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
Keith Moore writes: > I find myself wondering, don't they get support calls from customers > having to deal with the problems caused by the NATs? Sure, and the reply is "I'm sorry, but we don't support multiple computers on residential accounts." ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
RE: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and how to stop them.)
> From: Tim Chown [mailto:[EMAIL PROTECTED] > I noticed that by deafult MS Vista doesn't use autoconf as > per 2462, rather it uses a 3041-like random address. See: > http://www.microsoft.com/technet/itsolutions/network/evaluate/ > new_network.mspx This should hardly be a surprise. The inability of the IETF to accept as legitimate real world security concerns at the time 2462 was written was as notorious as its insistence on certain irrelevant concerns being absolute. It the real world I am actually very concerned if someone is able to determine my MAC address. It opens up a significant amount of information about my internal network that in the real world I have no intention of sharing. During the 1990s many people, myself included mistook applying cryptography for security. In the real world what network managers want to do and will insist on having tools to acomplish is to guard the boundary between the internal and external network as closely as possible and to prevent any piece of unnecessary information crossing that barrier in either direction. It is certainly a mistake to consider this practice a sufficient condition for security but anyonje who does not understand that it is a necessary party of a security strategy has little to contribute to today's security architecture. smime.p7s Description: S/MIME cryptographic signature ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was:
On Thu, Mar 30, 2006 at 07:08:51PM +0200, Anthony G. Atkielski wrote: > Steve Silverman writes: > > > The problem with allocating numbers sequentially is the impact on > > routers and routing protocols. > > The problem with not doing so is that a 128-bit address doesn't > provide anything even remotely close to 2^128 addresses. > But this has been known all along. It's a feature, not a bug. If we "throw away" the last 64 bits we are left with 2**64 addresses, which is obviously what was intended from the beginning. Allocating to ISPs on /48 or so boundaries then maps roughly to ipv4 /16s which are often handed out now. If a /8 in v4 space gets parceled out by the registries in a matter of months, a /40 in v6 should last the same. Following that logic, a /32 in v6 space should last as long as /0 in v4 space. The current v4 /0 has lasted for some time now, it's difficult to envision a time where we would be burning through space so fast that /32s in v6 space wouldn't last months, if not years. But for argument's sake, let's say a /32 lasts one day at some point in the dim future. This gives us 2 ** 32 days before we run out. not too bad for those of us not realistically having much of a chance to live beyond 2 ** 15 or so. If we "throw away" additional bits for other engineering purposes the number of days would be 2 ** (32 - wasted bits). If we waste a full half of those bits bits we're down to 2 ** 16 days (about 180 years). Again, that assumes we'd burn what is equivalent to a v4 /0 every single day for 180 years. Maybe it will become a problem when everyone has addressable nanocytes in their bloodstream. But I have a hard time seeing how it could be described as shortsighted. > In that case, assign addresses to points in space, instead of devices. > An office occupying a given plot of land will have an IP address space > that is solely a function of the space it occupies. Routing would be > the essence of simplicity and blazingly fast. > Routing doesn't (and will never) work that way. Much like with airlines, the cost of a path is more complex than mere distance. > > IMO one problem of the Internet is that it isn't hierarchical enough. > > Consider the phone system: country codes, area codes ... This makes > > the job of building a switch much easier. I think we should have > > divided the world into 250 countries. Each country into 250 > > "provinces". Yes, it would waste address space but it would make > > routing much easier and more deterministic. > > With a variable address length that can extend infinitely at either > end, the address space would never be exhausted. That's how > telephones work. > That would be an alternative, certainly. I'm not sure how excited to get about a 1 byte payload needing 1000 bytes of header, but I'm sure it's possible. Is it worth throwing away the current post-v4 solution? Certainly something to consider over the next 180 years. Austin ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
On Wed, Mar 29, 2006 at 01:00:44AM +0200, Iljitsch van Beijnum wrote: > 1996 199719981999200020012002200320042005 > 2.7 1.2 1.6 1.2 2.1 2.4 1.9 2.4 3.4 4.5 > > (The numbers represent the number of addresses used up in that year > as a percentage of the 3.7 billion total usable IPv4 addresses.) > > Those years where the growth was smaller than the year before never > happened twice or more in a row. > > This basically means that unless things take a radical turn, the long- > term trend is accelerating growth so that remaining 40% will be gone > in less than 9 years. Probably something like 7, as Geoff Huston > predicts. > This is much less time than I have seen in previous reports. If this is accurate and consistent there is a greater problem than I had previously thought. If that is indeed the case then the "enhanced nat" road for ipv6 begins to make much more sense, even in the nearer term. Austin ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
On 30-mrt-2006, at 22:13, Austin Schutz wrote: This basically means that unless things take a radical turn, the long- term trend is accelerating growth so that remaining 40% will be gone in less than 9 years. Probably something like 7, as Geoff Huston predicts. This is much less time than I have seen in previous reports. If this is accurate and consistent there is a greater problem than I had previously thought. You may want to look at Geoff's reports from 2003 and 2005 (and observe the difference between the two). Between 2003 and 2006 the growth was much larger than before 2003. As for accuracy: the numbers keep jumping up and down so you can pretty much massage them in either direction. These are the number of addresses the RIRs allocated to (mostly) ISPs per month for 2005 and 2006, compiled from the stats published on their FTP sites: 9.26 M 2005-01 25.16 M 2005-02 17.24 M 2005-03 24.96 M 2005-04 16.67 M 2005-05 16.12 M 2005-06 11.11 M 2005-07 16.92 M 2005-08 8.80 M 2005-09 9.89 M 2005-10 5.78 M 2005-11 6.62 M 2005-12 8.95 M 2006-01 11.88 M 2006-02 14.59 M 2006-03 There are now 1428 million addresses still free, give or take one or two /8s because of IANA/ARIN inconsistencies. So we can interpret this as: - maximum was 26.16 million/month, which gives us another 57 months - average was 13.6 million/month, which gives us another 105 months - average when ignoring bottom and top 3 values was 12.8, 111 months But that's without accelerating growth, which certainly has been the trend most of the time: since 1995, 7 out of 11 years saw more new IPv4 addresses deployed than the year before, 3 less. Unfortunately, the only way to discern a quantifiable trend here involves so much smoothing of the data that the results become meaningless, IMO. That said, with no growth in address usage, we'll be out in 2015 and since 1997, there has always been growth over any three year period. With _any_ growth we'll be out of IPv4 addresses in less than 9 years. It will take something as big as the deployment of CIDR to get us off that track. And no, NAT doesn't even come close. (We went from 211 M in 1991 to 63 M in 1994, since 1997 the biggest drop was about 25% in one year that was made up for the year after that both in 1999 and 2002.) There are plenty more interesting data points, such as the observation that even though China and India are similar in population, China has 77 million IPv4 addresses (just under the UK and just above Canada) and India 6 million (just under Denmark and just above Hong Kong). There is only one country with more addresses per capita than the US (~4): the Vatican (8192 addresses for 1000 inhabitants). And even though the US tops the absolute and per capita lists of address space holders, it still received the most addresses of any country in 2005: more than the numbers two (Japan) and three (China) together. If that is indeed the case then the "enhanced nat" road for ipv6 begins to make much more sense, even in the nearer term. I remember someone saying something about enhanced NAT here a few days ago but I can't find it... What is it and what does it have to do with IPv6? ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
Austin Schutz wrote: On Wed, Mar 29, 2006 at 01:00:44AM +0200, Iljitsch van Beijnum wrote: 1996199719981999200020012002200320042005 2.7 1.2 1.6 1.2 2.1 2.4 1.9 2.4 3.4 4.5 (The numbers represent the number of addresses used up in that year as a percentage of the 3.7 billion total usable IPv4 addresses.) Those years where the growth was smaller than the year before never happened twice or more in a row. This basically means that unless things take a radical turn, the long- term trend is accelerating growth so that remaining 40% will be gone in less than 9 years. Probably something like 7, as Geoff Huston predicts. This is much less time than I have seen in previous reports. If this is accurate and consistent there is a greater problem than I had previously thought. If that is indeed the case then the "enhanced nat" road for ipv6 begins to make much more sense, even in the nearer term. Austin I am afraid the problem is even bigger. I have seen again and again that cable providers are giving out ip-addresses in the 10.0.0.0/8 area to save ip address space. Not to mention wireless hotspots. The hotspots I have been playing with own only a single one ip-address. You notice something is awfully wrong, when your VoIP phone is not working but your neighbar keeps telling you, his skype does. -- Peter and Karin Dambier The Public-Root Consortium Graeffstrasse 14 D-64646 Heppenheim +49(6252)671-788 (Telekom) +49(179)108-3978 (O2 Genion) +49(6252)750-308 (VoIP: sipgate.de) mail: [EMAIL PROTECTED] mail: [EMAIL PROTECTED] http://iason.site.voila.fr/ https://sourceforge.net/projects/iason/ ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
--On Thursday, March 30, 2006 08:47 -0800 Peter Sherbin <[EMAIL PROTECTED]> wrote: If someone calls up for help with a configuration problem, that may be six month's of profits from that customer eaten up in the cost of answering the call. That is because the current Internet pricing has been screwed-up from the start. LD settlements between telcos are fully applicable to ISPs but have never been instituted. Internet has been subsidised for years by the local access but now as wireline declines everybody starts feeling the pain. Usage based billing and inter-ISP settlements start showing up lately and they fit well for the Internet. Otherwise transit providers as well as heavy users rip all the benefits. Peter, I was describing the facts of what is going on, rather than the causes. That said, based on some experience on both the telco and ISP sides of things, I believe your analysis is incorrect in a number of important ways, starting with the difficulties of applying a settlement model that assumes that value accrues to the caller to today's Internet. john ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
--On Thursday, March 30, 2006 11:44 -0500 Keith Moore wrote: However, we need to keep something else in mind, which Iljitsch's note hints at. If I'm an ISP trying to sell a low-end service to low-end customers at a low (but still profitable) price, I need to cut customer support costs to the absolute minimum. If someone calls up for help with a configuration problem, that may be six month's of profits from that customer eaten up in the cost of answering the call. I find myself wondering, don't they get support calls from customers having to deal with the problems caused by the NATs? Because they don't answer them. In the process of doing the work that led to RFC 4084, I reviewed the terms and conditions of service of a large number of ISPs in the US (and a few others) who provide low-cost Internet connectivity. Some prohibit connection of more than one machine to the incoming line/router/modem. Others provide a NAT-capable router but prohibit the customer from making any changes to its configuration and from running any applications that don't work in that environment. And still others indicate that customers can supply their own NATs, but must obtain any support elsewhere. All of these prohibitions are "enforced" the same way -- if the user calls with a problem, he or she either (i) is told that there is no support for violations of the rules and offered the opportunity to be disconnected (often with a large "early termination fee") or (ii) is instructed to disconnect all equipment between the machine in question and the router, and see if the problem still occurs. If it doesn't, then the ISP has no problem and the customer's problem is of no interest. For the community, there are elements of "you get what you pay for" in this. And, for the ISPs, unless we figure out ways to provide the same level of support convenience with public addresses, we will certainly see NATs with IPv6 as well as IPv4. either that, or IPv6 will be seen as something that is "business use only". That possibility had not occurred to me, but I fear you are correct. john ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
> > I find myself wondering, don't they get support calls from > > customers having to deal with the problems caused by the NATs? > > Because they don't answer them. In the process of doing the > work that led to RFC 4084, I reviewed the terms and conditions > of service of a large number of ISPs in the US (and a few > others) who provide low-cost Internet connectivity. Some > prohibit connection of more than one machine to the incoming > line/router/modem. Others provide a NAT-capable router but > prohibit the customer from making any changes to its > configuration and from running any applications that don't work > in that environment. And still others indicate that customers > can supply their own NATs, but must obtain any support > elsewhere. All of these prohibitions are "enforced" the same > way -- if the user calls with a problem, he or she either > > (i) is told that there is no support for violations of the rules > and offered the opportunity to be disconnected (often with a > large "early termination fee") or > > (ii) is instructed to disconnect all equipment between the > machine in question and the router, and see if the problem still > occurs. If it doesn't, then the ISP has no problem and the > customer's problem is of no interest. Well, the reason I asked is that when I got my DSL line, my ISP supplied me with a modem that does NAT - but only for a single host. As best as I can tell this is because the box needs to run PPPoE on the carrier side and DHCP on the host side, and the only way that the DHCP server can give the host an address under those conditions is to do NAT. So in this case (which I have no reason to believe is atypical) the ISP is supplying the NAT - and they do so even for customers who pay them extra to get a static IP address! And yes it does break things even when there are no other local hosts involved and no additional boxes between the modem and the customer's host. So I have a hard time believing that ISPs don't get support calls about failures due to NATs, at least when they install the NATs. Now of course this ISP does have a T&C that prohibits running a server, but "server" is a pretty vague term, and you don't have to be running any kind of server to suffer from NAT brain-damage. Keith p.s. fwiw the workaround in my case was to tell the modem to work in "passthrough" mode and configure my local router to run PPPoE. Under those conditions, I'm happy to report, 6to4 works just fine. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: Stupid NAT tricks and how to stop them.
On Thu, Mar 30, 2006 at 11:26:40PM +0200, Iljitsch van Beijnum wrote: > >If that is indeed the case then the "enhanced nat" road for ipv6 > >begins to make much more sense, even in the nearer term. > > I remember someone saying something about enhanced NAT here a few > days ago but I can't find it... What is it and what does it have to > do with IPv6? It was a term Keith Moore used to describe the addition of ipv6 capability to NAT devices. Not intended as a real term, merely a marketing name to explain to the end user the benefit of having ipv6 capability. If address space does indeed burn that quickly, ISPs will start to realize they can't sell additional IP addresses as a way of making a quick profit. Those with dwindling address pools will begin to demand proper ipv6 support from router vendors to offer it at a discounted price (compared to ipv4) to their customers who are savvy enough to want to run servers but too cheap to buy ipv4 space at a premium. From there it should only be a matter of time. If key applications work with ipv6 that will probably be adequate to get the ball rolling. IIRC there was a similar transition back when virtual web hosting meant blowing an ip address for every extra domain. After an adequate number of browsers were upgraded hosting providers made available ip-less virtual hosts at a heavy discount from ip-burning ones. After a surprisingly short amount of time the vast majority of browsers were compliant. The final nail was registries refusing virtual hosting as an excuse to justify allocations. That's not news to most here, but I definitely see the similarity in the situation. Austin ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was:
Austin Schutz writes: > But this has been known all along. It's a feature, not a bug. Yeah, right. > If we "throw away" the last 64 bits we are left with 2**64 addresses, > which is obviously what was intended from the beginning. And when you allocate bit fields in the remaining 64 bits, you exhaust that as well. > The current v4 /0 has lasted for some time now, it's difficult to > envision a time where we would be burning through space so fast that > /32s in v6 space wouldn't last months, if not years. The inability to envision things is the root of the problem. > But for argument's sake, let's say a /32 lasts one day at some point > in the dim future. This gives us 2 ** 32 days before we run out. not > too bad for those of us not realistically having much of a chance to > live beyond 2 ** 15 or so. More bogus math. Every time someone tries to compute capacity, he looks at the address space in terms of powers of two. Every time someone tries to allocate address space, he looks as the address space in terms of a string of bits. Since the space is allocated as bit fields, it is exhausted in linear time, even though capacity projections are based on an exponential space. For this reason, the address space always comes up too short, too soon. The real mystery is that this seems to surprise people who should know better. > If we "throw away" additional bits for other engineering purposes > the number of days would be 2 ** (32 - wasted bits). If we waste a full > half of those bits bits we're down to 2 ** 16 days (about 180 years). Yes, I know how these calculations work. See above. > Again, that assumes we'd burn what is equivalent to a v4 /0 > every single day for 180 years. It doesn't matter how fast we "burn" a v4 /0, because the space is exhausted by encoding information into the address and allocating bit spans in the address, not by actually handing out addresses. > Routing doesn't (and will never) work that way. Much like with > airlines, the cost of a path is more complex than mere distance. The cost of a processor will always be high and it will always fill three seven-foot-high cabinets, therefore all computers will always be timesharing systems. There's no need to imagine desktop computers; computing doesn't (and will never) work that way. We should prepare for the future of dumb terminals, which will eventually be everywhere (in every home and office). Do you see the problem with making predictions about the future? > That would be an alternative, certainly. I'm not sure how excited > to get about a 1 byte payload needing 1000 bytes of header, but I'm sure > it's possible. It would only need a thousand bytes if it were being routed to the other side of the universe. > Is it worth throwing away the current post-v4 solution? Throwing away a few years' work for something that makes virtually no assumptions about the future? Yes, it might well be worth it. It would have been worth it if it had been done that way in the first place, too. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and how to stop them.)
Thus spake "Anthony G. Atkielski" <[EMAIL PROTECTED]> Iljitsch van Beijnum writes: However, since that time I've learned to appreciate stateless autoconfiguration and the potential usefulness of having the lower 64 bits of the IPv6 address as a place to carry some limited security information (see SEND and shim6 HBA). Once it's carrying information, it's no longer just an address, so counting it as pure address space is dangerous. An IPv4/6 address is both a routing locator and an interface identifier. Unfortunately, the v6 architects decided not to separate these into separate address spaces, so an address _must_ contain routing information until that problem is fixed. It doesn't seem to be likely we'll do so without having to replace IPv6 and/or BGP4+, and there's no motion on either front, so we're stuck with the locator/identifier problem for quite a while. Building in space means not allocating it--not even _planning_ to allocate it. Nobody has any idea what the Internet might be like a hundred years from now, so why are so many people hellbent on "planning" for something they can't even imagine? That's why 85% of the address space is reserved. The /3 we are using (and even then only a tiny fraction thereof) will last a long, long time even with the most pessimistic projections. If it turns out we're still wrong about that, we can come up with a different policy for the next /3 we use. Or we could change the policy for the existing /3(s) to avoid needing to consume new ones. If IPv6 is supposed to last 100 years, that means we have ~12.5 years to burn through each /3, most likely using progressively stricter policies. It's been a decade since we started and we're nowhere near using up the first /3 yet, so it appears we're in no danger at this point. Will we be in 50 years? None of us know, which is why we've reserved space for the folks running the Internet then to make use of -- provided IPv6 hasn't been replaced by then and making this whole debate moot. Unfortunately, at the time IPv6 was created variable length addresses weren't considered viable. Variable-length addresses are the only permanent solution, unless IP addresses are assigned serially (meaning that all routing information has to be removed). Variable-length addresses work very well for the telephone system, and they'd work just as well for the Internet, if only someone had taken the time to work it out. Variable-length addresses only work if there is no maximum length. E.164 has a maximum of 15 digits, meaning there are at most 10^15 numbers. Here in +1 we only use eleven digit numbers, meaning we're burning them 10^4 times as fast as we could. That's not a great endorsement. Also, telephone numbers have the same locator/identifier problem that IPv4/6 addresses do. In fact, IPv6's original addressing model looked strikingly similar to the country codes and area/city codes (aka TLAs and NLAs) that you're apparently fond of. Even OSI's "variable length" addresses had a maximum length, and most deployments used the maximum length; they degenerated into fixed-length addresses almost immediately. The only thing I'm not too happy about is the current one address / one subnet / /48 trichotomy. Ignoring the single address for a moment, the choice between one subnet and 65536 isn't a great one, as many things require a number of subnets that's greater than one, but not by much. It's a good example of waste that results from short-sightedness. It happened in IPv4, too. The difference is that in IPv6, it's merely a convention and implementors are explicitly told that they must not assume the above boundaries. In IPv4, it was hardcoded into the protocol and every implementation had to be replaced to move to VLSM and CIDR. Conventions are for human benefit, but they can be dropped when it becomes necessary. Folks who use RFC 1918 space almost always assign /24s for each subnet regardless of the number of hosts; folks using public addresses used to do the same, but instead now determine the minimum subnet that meets their needs. Hopefully the conventions in IPv6 won't be under attack for a long time, but if they need to go one day we can drop them easily enough. The thing that is good about IPv6 is that once you get yourself a / 64, you can subdivide it yourself and still have four billion times the IPv4 address space. It sounds like NAT. Not at all. You'd still have one address per host, you'd just move the subnet boundary over a few bits as needed. With the apparent move to random IIDs, there's no reason to stick to /64s for subnets -- we could go to /96s for subnets without any noticeable problems (including NAT). The RIRs could also change policies so that /32 and /48 are not the default allocation and assignment sizes, respectively. That is also another convention that we could easily dispense with, but it saves us a lot of paperwork to abide by i
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and how to stop them.)
Thus spake "Anthony G. Atkielski" <[EMAIL PROTECTED]> Iljitsch van Beijnum writes: So how big would you like addresses to be, then? It's not how big they are, it's how they are allocated. And they are allocated very poorly, even recklessly, which is why they run out so quickly. It's true that engineers always underestimate required capacity, but 128-bit addresses would be enough for anything ... IF they were fully allocated. But I know they won't be, and so the address space will be exhausted soon enough. I once read that engineers are generally incapable of designing anything that will last (without significant redesign) beyond their lifespan. Consider the original NANP and how it ran out of area codes and exchanges around 40 years after its design -- roughly the same timeframe as the expected death of its designers. Will IPv6 last even that long? IMHO we'll find a reason to replace it long before we run out of addresses, even at the current wasteful allocation rates. We currently have 1/8th of the IPv6 address space set aside for global unicast purposes ... Do you know how many addresses that is? One eighth of 128 bits is a 125-bit address space, or 42,535,295,865,117,307,932,921,825,928,971,026,432 addresses. That's enough to assign 735 IP addresses to every cubic centimetre in the currently observable universe (yes, I calculated it). Am I the only person who sees the absurdity of wasting addresses this way? It doesn't matter how many bits you put in an address, if you assign them this carelessly. That's one way of looking at it. The other is that even with just the currently allocated space, we can have 35,184,372,088,832 sites of 65,536 subnets of 18,446,744,073,709,551,616 hosts. Is this wasteful? Sure. Is it even conceivable to someone alive today how we could possibly run out of addresses? No. Will someone 25 years from now reach the same conclusion? Perhaps, perhaps not. That's why we're leaving the majority of the address space reserve for them to use in light of future requirements. ... with the idea that ISPs give their customers /48 blocks. Thank you for illustrating the classic engineer's mistake. Stop thinking in terms of _bits_, and think in terms of the _actual number of addresses_ available. Of better still, start thinking in terms of the _number of addresses you throw away_ each time you set aside entire bit spans in the address for any predetermined purpose. Remember, trying to encode information in the address (which is what you are doing when you reserve bit spans) results in exponential (read incomprehensibly huge) reductions in the number of available addresses. It's trivially easy to exhaust the entire address space this way. If you want exponential capacity from an address space, you have to assign the addresses consecutively and serially out of that address space. You cannot encode information in the address. You cannot divided the address in a linear way based on the bits it contains and still claim to have the benefits of the exponential number of addresses for which it supposedly provides. Why is this so difficult for people to understand? And sequential assignments become pointless even with 32-bit addresses because our routing infrastructure can't possibly handle the demands of such an allocation policy. The IETF has made the decision to leave the current routing infrastructure in place, and that necessitates a bitwise allocation model. Railing against this decision is pointless unless you have a new routing paradigm ready to deploy that can handle the demands of a non-bitwise allocation model. Why is this so difficult for you to understand? That gives us 45 bits worth of address space to use up. You're doing it again. It's not 45 bits; it's a factor of 35,184,372,088,832. But rest assured, they'll be gone in the blink of an eye if the address space continues to be mismanaged in this way. I take it you mean "the blick of an eye" to mean a span of decades? That is not the common understanding of the term, yet that's how long we've been using the current system and it shows absolutely no signs of strain. It's generally accepted that an HD ratio of 80% should be reachable without trouble, which means we get to waste 20% of those bits in aggregation hierarchies. No. It's not 20% of the bits, it's 99.9756% of your address space that you are wasting. Do engineers really study math? To achieve bitwise aggregation, you necessarily cannot achieve better than 50% use on each delegation boundary. There are currently three boundaries (RIR, LIR, site), so better than 12.5% address usage is a lofty goal. Again, if you want something better than this, you need to come up with a better routing model than what we have today. (And then throw in the /64 per subnet and you're effectively wasting 100% of the address space anyways, so none of this matters until that's gone) This gives us 36 bits = 6
Re: Stupid NAT tricks and how to stop them.
Thus spake "Keith Moore" Now of course this ISP does have a T&C that prohibits running a server, but "server" is a pretty vague term, and you don't have to be running any kind of server to suffer from NAT brain-damage. My ISP has ingeniously defined a "server" as any application that does not work through NAT without port forwarding. Bingo, problem solved (from their perspective). Of course, they don't actually enforce this unless a user's upstream bandwidth usage consistently exceeds total POP upstream bandwidth divided by the number of users at the POP (in my case, about 300kB/s). Go above that and you get an email asking you to turn down the speed on your P2P client ;-) p.s. fwiw the workaround in my case was to tell the modem to work in "passthrough" mode and configure my local router to run PPPoE. Under those conditions, I'm happy to report, 6to4 works just fine. Alas, I've been unable to find a consumer-grade router that will run native IPv6, 6to4, or even pass through IPinIP (excluding open-source hacks which are not supported by the vendor -- that's not a solution for real consumers). If anyone knows of one, please let me know off-list. S Stephen Sprunk"Stupid people surround themselves with smart CCIE #3723 people. Smart people surround themselves with K5SSS smart people who disagree with them." --Aaron Sorkin ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
RE: PI space (was: Stupid NAT tricks and how to stop them)
> Noel Chiappa wrote: > Needless to say, the real-time taken for this process to complete > - i.e. for routes to a particular destination to stabilize, after > a topology change which affects some subset of them - is dominated > by the speed-of-light transmission delays across the Internet > fabric. You can make the speed of your processors infinite and it > won't make much of a difference. This is total bull. The past stability issues in BGP have little to do with latency and everything to do with processing power and bandwidth available to propagate updates. In other words, it does not make any difference in the real world if you're using a 150ms oceanic cable or a 800ms geosynchronous satlink as long as the pipe is big enough and there are enough horses under the hood. Only if we were shooting for a sub-second global BGP convergence the speed of light would matter. > Stephen Sprunk wrote: > The IPv4 core is running around 180k routes today, and even > the chicken littles aren't complaining the sky is falling. I was about to make the same point. Ever heard whining about a 7500 with RSP2s not being able to handle it? Yes. Ever heard about a decently configured GSR not being able to handle it? No. Heard whining about receiving a full table over a T1? Yes. Heard whining about receiving a full table over an OC-48? No. Anybody still filtering at /20 like everybody did a few years back? > and the vendors can easily raise those limits if customers demand > it (though they'd much prefer charging $1000 for $1 worth of RAM > that's too old to work in a modern PC). You're slightly exaggerating here. I remember paying $1,900 for 32MB of ram worth $50 in the street :-D Michel. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
RE: PI space (was: Stupid NAT tricks and how to stop them)
> From: "Michel Py" <[EMAIL PROTECTED]> >> Needless to say, the real-time taken for this process to complete >> - i.e. for routes to a particular destination to stabilize, after a >> topology change which affects some subset of them - is dominated by >> the speed-of-light transmission delays across the Internet fabric. You >> can make the speed of your processors infinite and it wqwon't make >> much of a difference. > The past stability issues in BGP have little to do with latency and > everything to do with processing power and bandwidth available to > propagate updates. The past stability issues had a number of causes, including protocol implementation issues, IIRC. In any event, I was speaking of the present/future, not the past. Yes, *in the past*, processing power and bandwidth limits were an *additional* issue. However, that was in the past - *now*, the principal term in stabilization time is propogation delay. > In other words, it does not make any difference in the real world if > you're using a 150ms oceanic cable or a 800ms geosynchronous satlink as > long as the pipe is big enough and there are enough horses under the > hood. If you think there aren't still stability issues, why don't you try getting rid of all the BGP dampening stuff, then? Have any major ISP's out there done that? Noel ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and how to stop them.)
On Thu, 30 Mar 2006 20:43:14 -0600, "Stephen Sprunk" <[EMAIL PROTECTED]> wrote: > > That's why 85% of the address space is reserved. The /3 we are using (and > even then only a tiny fraction thereof) will last a long, long time even > with the most pessimistic projections. If it turns out we're still wrong > about that, we can come up with a different policy for the next /3 we use. > Or we could change the policy for the existing /3(s) to avoid needing to > consume new ones. > I really shouldn't waste my time on this thread; I really do know better. You're absolutely right about the /3 business -- this was a very deliberate design decision. So, by the way, was the decision to use 128-bit, fixed-length addresses -- we really did think about this stuff, way back when. When the IPng directorate was designing/selecting what's now IPv6, there was a variable-length address candidate on the table: CLNP. It was strongly favored by some because of the flexibility; others pointed out how slow that would be, especially in hardware. There was another proposal, one that was almost adopted, for something very much like today's IPv6 but with 64/128/192/256-bit addresses, controlled by the high-order two bits. That looked fast enough in hardware, albeit with the the destination address coming first in the packet. OTOH, that would have slowed down source address checking (think BCP 38), so maybe it wasn't a great idea. There was enough opposition to that scheme that a compromise was reached -- those who favored the 64/128/192/256 scheme would accept fixed-length addresses if the length was changed to 128 bits from 64, partially for future-proofing and partially for flexibility in usage. That decision was derided because it seemed to be too much address space to some, space we'd never use. I'm carefully not saying which option I supported. I now think, though, that 128 bits has worked well. --Steven M. Bellovin, http://www.cs.columbia.edu/~smb ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
RE: Stupid NAT tricks and how to stop them.
> Noel Chiappa wrote: > If you think there aren't still stability issues, why don't > you try getting rid of all the BGP dampening stuff, then? > Have any major ISP's out there done that? Dampening is part of the protocol and has nothing to do with the speed of light. Removing it is akin to removing packet re-ordering in TCP; nobody with half a brain would consider it. Same as packets arriving out of sequence, stability issues are part of life for someone who actually operates a network. Code has bugs, hardware fails, power goes bezerk, UPS batteries leak, rodents chew on cables, backhoes cut fiber, fat fingers screw up configs, and rookies flap routes because they know everything about astrophysics and nothing about running a production network. That's what dampening is for. > Stephen Sprunk wrote: > Of course, they don't actually enforce this unless a user's > upstream bandwidth usage consistently exceeds total POP > upstream bandwidth divided by the number of users at the POP > (in my case, about 300kB/s). Go above that and you get an email > asking you to turn down the speed on your P2P client ;-) Some bigger ISPs (with large POPs) prefer to limit the upstream so they don't have to manage quotas/bandwidth. I have 512kb upstream; it's not big but I can bittorrent all of it all day long, they don't care and probably don't know. > Alas, I've been unable to find a consumer-grade router that > will run native IPv6, 6to4, or even pass through IPinIP There's no market for it. Consumers don't know what it is and geeks already have a hack or an el-cheapo-ebay Cisco. Heck, my home router is a 7204 why would I go for a Linksys :-D Michel. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was:
On Fri, Mar 31, 2006 at 05:36:30AM +0200, Anthony G. Atkielski wrote: > More bogus math. Every time someone tries to compute capacity, he > looks at the address space in terms of powers of two. Every time > someone tries to allocate address space, he looks as the address space > in terms of a string of bits. Anthony, You've been making the same point over and over (and over) again. It's probably the case that people who will be convinced by your arguments, will have accepted the force of your arguments by now. For people who don't accept your arguments, they are not likely to be swayed by a "last post wins" style of argumentation. May I gently suggest that you stop and think before deciding whether you need to respond to each message on this thread, and whether you have something new and cogent to add, as opposed to something which you've said already, in some cases multiple times? Thanks, regards, - Ted ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and how to stop them.)
Stephen Sprunk writes: > And sequential assignments become pointless even with 32-bit > addresses because our routing infrastructure can't possibly handle > the demands of such an allocation policy. They are pointless for the reasons you state, but they are also the only way to get 2^128 addresses out of 128 bits. Anything else encodes information in the address and reduces the usable address space exponentially. > Railing against this decision is pointless unless you have a new > routing paradigm ready to deploy that can handle the demands of a > non-bitwise allocation model. The bitwise allocations I'm hearing about are not based on routing. > I take it you mean "the blick of an eye" to mean a span of decades? At best. > That is not the common understanding of the term, yet that's how > long we've been using the current system and it shows absolutely no > signs of strain. So IPv6 is not needed? > To achieve bitwise aggregation, you necessarily cannot achieve > better than 50% use on each delegation boundary. There are currently > three boundaries (RIR, LIR, site), so better than 12.5% address > usage is a lofty goal. Again, if you want something better than > this, you need to come up with a better routing model than what we > have today. I did, but nobody was interested. > Again, the current identifier/location conflation combined with the > routing paradigm leaves us no choice but to encode information into > the IP address. In that case, any predictions of longevity for the system based on the address space providing 2^n addresses for n bits are invalid. Strangely, such predictions seem to be almost exclusively based on this, and thus are necessarily wrong. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
RE: Stupid NAT tricks and how to stop them.
> Dampening is part of the protocol and has nothing to do with the speed > of light. Well, not really. Assume a simplistic model of the Internet with M "core" routers (in the default free zone) and N "leaf" AS, i.e. networks that have their own non-aggregated prefix. Now, assume that each of the leaf AS has a "routing event" with a basic frequency, F. Without dampening, each core router would see each of these events with that same frequency, F. Each router would thus see O(N*F) events per second. Since events imply some overhead in processing, message passing, etc, one can assume that at any given point in time there is a limit to what a router can swallow. In either N or F is too large, the router is cooked. Hence dampening at a rate D, so that N*F/D remains lower than the acceptable limit. Bottom line, you can only increase the number of routes if you are ready to dampen more aggressively. There is an obvious "tragedy of the commons" here: if more network want to "multi-home" and be declared in the core, then more aggressive dampening will be required, and each of the "multi-homed" networks will suffer from less precise routing, longer time to correct outages, etc. There are different elements at play that also limit the number of core routers. Basically, an event in a core router affects all the path that go through it, which depending on the structure of the graph is somewhere between O(M*log(M)) or O(M.log(M)). In short, the routing load grows much faster than linearly with the number of core routers. -- Christian Huitema ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was: IPv6 vs. Stupid NAT tricks: false dichotomy? (Was: Re: StupidNAT tricks and how to stop them.)
Stephen Sprunk writes: > An IPv4/6 address is both a routing locator and an interface identifier. And so engineers should stop saying that n bits of addressing provides 2^n addresses, because that is never true if any information is encoded into the address. In fact, as soon as any information is placed into the address itself, the total address space shrinks exponentially. > Unfortunately, the v6 architects decided not to separate these into > separate address spaces, so an address _must_ contain routing > information until that problem is fixed. It doesn't seem to be > likely we'll do so without having to replace IPv6 and/or BGP4+, and > there's no motion on either front, so we're stuck with the > locator/identifier problem for quite a while. Then we need to make predictions for the longevity of the scheme based on the exponentially reduced address space imposed by encoding information into the address. In other words, 128 bits does _not_ provide 2^128 addresses; it does not even come close. Ultimately, it will barely provide anything more than what IPv4 provides, if current trends continue. > That's why 85% of the address space is reserved. The /3 we are using (and > even then only a tiny fraction thereof) will last a long, long time even > with the most pessimistic projections. If it turns out we're still wrong > about that, we can come up with a different policy for the next /3 we use. > Or we could change the policy for the existing /3(s) to avoid needing to > consume new ones. Or simply stop trying to define policies for an unknown future, and thereby avoid all these problems to begin with. > It's been a decade since we started and we're nowhere near using up the > first /3 yet, so it appears we're in no danger at this point. As soon as you chop off 64 bits for another field, you've lost just under 100% of it. > Variable-length addresses only work if there is no maximum length. Ultimately, yes. But there is no reason why a maximum length must be imposed. > E.164 has a maximum of 15 digits, meaning there are at most 10^15 > numbers. Here in +1 we only use eleven digit numbers, meaning we're > burning them 10^4 times as fast as we could. That's not a great > endorsement. Telephone engineers make the same mistakes as anyone else; no natural physical law imposes E.164, however. > Also, telephone numbers have the same locator/identifier problem > that IPv4/6 addresses do. In fact, IPv6's original addressing model > looked strikingly similar to the country codes and area/city codes > (aka TLAs and NLAs) that you're apparently fond of. Maybe the problem is in trying to make addresses do both. Nobody tries to identify General Electric by its street address, and nobody tries to obtain a street address based on the identifier General Electric alone. > The difference is that in IPv6, it's merely a convention ... Conventions cripple society in many cases, so "merely a convention" may be almost an oxymoron. > The folks who designed IPv4 definitely suffered from that problem. The > folks who designed IPv6 might also have suffered from it, but at least they > were aware of that chance and did their best to mitigate it. Could they > have done better? It's always possible to second-guess someone ten years > later. There's also plenty of time to fix it if we develop consensus > there's a problem. Sometimes the most important design criterion is ignorance. In other words, the best thing an engineer can say to himself in certain aspects of design is "I don't know." ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Re: 128 bits should be enough for everyone, was:
Theodore Ts'o writes: > You've been making the same point over and over (and over) > again. To some, perhaps. I'm not so sure that it has yet been made even once to others. > It's probably the case that people who will be convinced by your > arguments, will have accepted the force of your arguments by now. > For people who don't accept your arguments, they are not likely to > be swayed by a "last post wins" style of argumentation. It depends. People with an emotional attachment to a specific notion will never been convinced otherwise, but people who simply don't understand something may change their mind once they understand. > May I gently suggest that you stop and think before deciding > whether you need to respond to each message on this thread, and > whether you have something new and cogent to add, as opposed to > something which you've said already, in some cases multiple times? May I gently suggest that you use the delete key on your keyboard for messages that you don't want to see? I doubt that bandwidth is a problem at MIT. It has always worked for me, and I'm constrained by much more limited bandwidth. ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf
Don't feed the trolls
On Thu, 30 Mar 2006, Stephen Sprunk wrote: Thus spake "Anthony G. Atkielski" <[EMAIL PROTECTED]> Why is this so difficult for people to understand? .. Why is this so difficult for you to understand? Feeding trolls is not very useful.. please at least keep him in Cc: so I don't have to see the messages.. :-) I've yet to see Anthony make any useful contribution to the IETF (various rants on the IETF list don't count). Perhaps we should just ban him from the list. -- Pekka Savola "You each name yourselves king, yet the Netcore Oykingdom bleeds." Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings ___ Ietf mailing list Ietf@ietf.org https://www1.ietf.org/mailman/listinfo/ietf