Does anyone but me agree that the Internic's current address allocation policy is counter-productive? I've been trying for three weeks now to get a block of addresses assigned to me for re-assignment to my customers. I run the non-profit Internet in the whole state of Iowa and the Internic asks me to tell them *ahead of time* how many hosts there will be and the subnet and masking policy for this block. How the H*** am I supposed to know that? Since I sent in this request, I have had *legitimate* requests for over 40 class C-sized blocks. If I have to go to my regional providers block to satisfy them, it will just contribute to the global routing table explosion. My regional provider, Midnet, tells me that to get the last CIDR block, they had to put in over 16 man-hours convincing the Internic that their request was valid. This is from a Regional that serves 7 states! Is this crazy, or what? Now, what do we do about it? Paul Lustgraaf "Its easier to apologize than to get permission." Network Specialist Grace Hopper Iowa State University Computation Center grpjl@iastate.edu Ames, IA 50011 515-294-0324
YES. Setting up the SWIP and some other requirements might be expected but, as providers I think the line needs to be drawn somewhere. I mean there is a difference between the smaller isp and a network thats got millions of dollars invested. I know that other major nets have had problems as well. This problem just needs to go away. Joseph Stroup On Fri, 17 Mar 1995, Paul Lustgraaf wrote:
Does anyone but me agree that the Internic's current address allocation policy is counter-productive? I've been trying for three weeks now to get a block of addresses assigned to me for re-assignment to my customers. I run the non-profit Internet in the whole state of Iowa and the Internic asks me to tell them *ahead of time* how many hosts there will be and the subnet and masking policy for this block.
How the H*** am I supposed to know that?
Since I sent in this request, I have had *legitimate* requests for over 40 class C-sized blocks. If I have to go to my regional providers block to satisfy them, it will just contribute to the global routing table explosion.
My regional provider, Midnet, tells me that to get the last CIDR block, they had to put in over 16 man-hours convincing the Internic that their request was valid. This is from a Regional that serves 7 states!
Is this crazy, or what?
Now, what do we do about it?
Paul Lustgraaf "Its easier to apologize than to get permission." Network Specialist Grace Hopper Iowa State University Computation Center grpjl@iastate.edu Ames, IA 50011 515-294-0324
You know the more I read on this the nutter it sounds. We only get address numbers to use them ? I sure as heck don't see having a Class A, B, C or anything like that as a status symbol. Ip numbers are like a Craftsman wrench, a means to and end. How many others are having this same problem ? Joseph
It would be a good idea to recover the ip address numbers that have never been used. I would also suggest that within 90 days of issuance the numbers must be in service or they are recalled. I mentioned to the Nic that there are several class B numbers out there being used as a "status symbol" , not for the purpose stated. All I hear is lets talk about not and not the past. For international X.25 services NET-99 required a DNIC number. The FCC inserts a clause states the following: all DNIC assignments are provisional and no assignee obtains an ownership or property interest in a DNIC assignment. As a rseult, the DNIC herein assigned is expressly subject to possible revocation or reassignment as may be required because of a shortage of DNIC's or as otherwise required in the public interest. If you fail to agree, you don't get a number. Throughout the FCC letters there is a constant mention of "this scarce DNIC resource may necessitate reassignment of codes not implemented within a six-month period. Now if the FCC has been doing this I am sure we could work out something with the Nic. Joseph Stroup
On Sat, 18 Mar 1995, ATM_Feel_the_Power wrote:
It would be a good idea to recover the ip address numbers that have never been used. I would also suggest that within 90 days of issuance the numbers must be in service or they are recalled.
I agree that unused address _blocks_ be recovered. However your 90 day "use-it-or-lose-it" is a bit rash. From a fellow-provider standpoint, I want to be able to get a nice-sized block of "class-c" addresses that I can aggregate into _1_ routing statement. To do that, I have to have the freedom to project client usage for a year and request a year's supply. Your suggestion would have me applying to the NIC on a "just-in-time" basis, and creating at least one more routing entry for each new client. Backbone routers are melting down already because of too many entries. And it is not like the NIC has the time to take on the more frequent requests for addresses and keep track of who's time is up that your plan would impose (or if they do, I want to know where a couple of domain registrations are :) ). My vote would be for the six-twelve month timeframe for useage. -- Michael Ramsey (KD4OKR) Network Engineering | INTERPATH msr@interpath.net | info@interpath.net (919)890-6300v (919)890-6319f | http://www.interpath.net 711 Hillsborough St., Raleigh, NC 27605 | helpdesk@interpath.net
On Sat, 18 Mar 1995, ATM_Feel_the_Power wrote:
It would be a good idea to recover the ip address numbers that have never been used. I would also suggest that within 90 days of issuance the numbers must be in service or they are recalled.
I agree that unused address _blocks_ be recovered.
However your 90 day "use-it-or-lose-it" is a bit rash. From a fellow-provider standpoint, I want to be able to get a nice-sized block of "class-c" addresses that I can aggregate into _1_ routing statement. To do that, I have to have the freedom to project client usage for a year and request a year's supply. Your suggestion would have me applying to the NIC on a "just-in-time" basis, and creating at least one more routing entry for each new client. Backbone routers are melting down already because of too many entries. And it is not like the NIC has the time to take on the more frequent requests for addresses and keep track of who's time is up that your plan would impose (or if they do, I want to know where a couple of domain registrations are :) ).
My vote would be for the six-twelve month timeframe for useage.
-- Michael Ramsey (KD4OKR) Network Engineering | INTERPATH msr@interpath.net | info@interpath.net
Of course, when we projected that, they didn't believe us -- and refused to grant the block. My request was for 256 Class "C"s. We have consumed almost exactly that in a year. Yet, we were told to go stuff a year ago when we asked for the block, and instead have had to get four different smaller blocks of space from other sources. And we are not alone; I am personally involved with another NATIONAL provider who has been told to stuff more than once by these folks, and they consume addresses rather more quickly than we do. This sucks, and in addition, it is counter-productive. Announcing one aggregate beats announcing 4. Of course, the NIC does not now and never has in the past given a damn about getting ISPs *appropriate* sized blocks of addresses, nor does it believe you when you call or email and request an allocation. I've been told several times "the evidence does not support that allocation" and then turned around and exceeded my own estimates (which I tend to be conservative with) within a few months. Perhaps its time to set up the Anti-NIC and start allocating addresses from 223.255.255.x on down. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
. . .
Perhaps its time to set up the Anti-NIC and start allocating addresses from 223.255.255.x on down.
If it weren't so sad, I would find this whole exchange very entertaining. I doubt that creating an anarchy will solve the problem. In the end, the InterNIC is a contractor, and does what they are being told. It is a service provided by the Feds for the better of the community. I am not trying to defend the InterNIC, just saying that they are more or less a robot operating on some guidelines. If you have a robot with better guidelines, more power to you. I suspect the problem is not the robot, but the guidelines (the program it executes). May be the guidelines should be worked on, and I encourage the people who complain on this list to develop better guidelines, then get community consensus, then go to the funding agenci(es) that are responsible for the NICs and present your case to them. If you get that far, I suspect your likelihood of success to achieving a change is above the 95th percentile. I know that both the InterNIC as well as Jon Postel have explicitly asked for guidance from constituents over the years. Seems to me like they have been left alone and now people are complaining, given the reaction on both sites of the addressing space being a scarce resource. Of course, I cannot help the thought that the community could have chosen (extendable) NSAPs years back as part of a CLNP package. Hans-Werner
Perhaps its time to set up the Anti-NIC and start allocating addresses from 223.255.255.x on down.
If it weren't so sad, I would find this whole exchange very entertaining.
I doubt that creating an anarchy will solve the problem. In the end, the InterNIC is a contractor, and does what they are being told. It is a service provided by the Feds for the better of the community. I am not trying to defend the InterNIC, just saying that they are more or less a robot operating on some guidelines. If you have a robot with better guidelines, more power to you. I suspect the problem is not the robot, but the guidelines (the program it executes). May be the guidelines should be worked on, and I encourage the people who complain on this list to develop better guidelines, then get community consensus, then go to the funding agenci(es) that are responsible for the NICs and present your case to them.
The problem here is that the NIC doesn't have any "authority" per-se. They may have TAKEN that authority, but it didn't evolve from anything coherent. The US Government, which funded this enterprise, doesn't have "control" over the address and namespace. They might think they do, but the fact is that nowhere did the International community grant them that, and this is NOT a US resource to delegate as it sees fit. The Internic's registrations are only honored because the ISPs and consumers of Internet service, in the main, believe the information that the Internic gives out and contains. If that EVER ceases to be, then the Internic can do whatever the hell it wants, and the net effect will be zero.
If you get that far, I suspect your likelihood of success to achieving a change is above the 95th percentile. I know that both the InterNIC as well as Jon Postel have explicitly asked for guidance from constituents over the years. Seems to me like they have been left alone and now people are complaining, given the reaction on both sites of the addressing space being a scarce resource.
Hans-Werner
Nonsense. We have not left any of them alone; I have frequently posited positive changes and suggestions. The entire top-level domain structure is a botch, and geographic registration is increasingly useless with mobile people and multi-state (or multinational) corporations. Why not have the Internic do two things only: 1) Issue top level domains to registries, anyone may apply. (Ie: countries which wish to have an official "top level" may apply, etc) 2) Issue LARGE blocks of addresses to those same registries, subject only to the requirement that you issue 75% of the addresses you get before you get more. Delegating an address to the final consumer or intermediary should be a 15-second process. 3) Charge a modest fee, say, US $1,000/year, for the privilege of getting (1) and (2) from the Internic, with said funds to be used to pay the ACTUAL costs of maintaining registries for (1) and (2). This resolves a number of problems at once: 1) Address allocation -- if a registry wants to charge, fine. This is now a free market solution with free market prices. I suspect you'd see a modest annual charge show up immediately from most registries -- which would stem the issuance and retention of addresses which are not actually in use. 2) Domain "fights". No longer is there one true registry. Thus, the entire "my company has a trademark on the name frobozz" disappears; there can now be a "frobozz.psi", "frobozz.mcs", "frobozz.us", and "frobozz.co.ac". All distinct, and yet all under the same name. Whois needs to be able to deal with this, which means a distributed system. This is difficult? At the ABSOLUTE worst case we have bounded the damages for a "sorry, can't register" response -- its now $1,000/year, and not worth litigating over (you can set up your OWN registry for that cost). 3) The responsiveness issue. Domains take 3-6 weeks to get back now. That's rediculous. Registries could compete based on cost and performance; now, if you want to register something right NOW, you need to beg, and even then, the Internic's idea of "now" is a couple of days. But with the "you pay to play" multiple-registry system, I could offer 10 day turnaround for $5, and "right now damnit" turnaround for $500. As things stand at present I can't offer anything, as I can't get a top-level domain to have administrative authority over. 4) The Internic is no longer a government thing. Rather, it is a non-profit funded BY THE USERS and accountable TO THE USERS -- that is, the operators of the ACTUAL end-user registries. What a concept we have here -- accountability to the people paying the bill! Its about time the Internic was privatized. The function is VERY important, but we've got a system which is too centralized right now and not accountable to anyone. This has to stop. At the same time a for-profit motive is not acceptable; this is a community resource, and as such a non-profit structure accountable to the membership -- with a SMALL charge, paid by all who use it directly, appears appropriate for the top level registration authority. If the harassment, and yes, that is what I consider being asked to submit BUSINESS PLANS to a government agency in order to get address space, continues then the solution to this is going to be found either (1) in the court system, or (2) by a group of people setting up their own "Internic" and saying "bite me" to the existing one. The first is nasty -- the second, which appearing to be an anarchist's solution, may actually be the best and only path out of this mess considering the parochial attitudes I have run into at the Internic itself. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
Well, I see two problems, one is that whatever solution be found needs community consensus. Otherwise we may end up with ten anti-NICs or so? Or at least many unhappy customers? The other problem is that (at least I, but may be I am missing something) don't see how at the global systems level this resolves the issue of appropriating address space to, say, a j-random, perhaps new and ignorant with uncertain future, service provider who would like to have 10,000 customers a year from now, and wants 10,000 Class-C numbers accordingly. I still don't understand what the groundrules for address allocation would be, at least at the level below your top-level allocations. Even if you push the problem a level down, it would still have to be resolved. The resource will be scarce somewhere. I am trying to focus on the IP address allocation guidelines here, not the speediness of registrations, with the latter being more of a technical (though apparently painful) issue, and probably a matter of having enough resources allocated.
Well, I see two problems, one is that whatever solution be found needs community consensus. Otherwise we may end up with ten anti-NICs or so? Or at least many unhappy customers? The other problem is that (at least I, but may be I am missing something) don't see how at the global systems level this resolves the issue of appropriating address space to, say, a j-random, perhaps new and ignorant with uncertain future, service provider who would like to have 10,000 customers a year from now, and wants 10,000 Class-C numbers accordingly. I still don't understand what the groundrules for address allocation would be, at least at the level below your top-level allocations. Even if you push the problem a level down, it would still have to be resolved. The resource will be scarce somewhere.
I am trying to focus on the IP address allocation guidelines here, not the speediness of registrations, with the latter being more of a technical (though apparently painful) issue, and probably a matter of having enough resources allocated.
Well, sure, the new ISP might want 10,000 Class "C" addresses. But the new ISP has no basis to request these. However, the CURRENT NIC is declining requests for *256* Class "C" addresses! That is, I was turned down for a Class "B" equivalent in a CIDR block. That is ludicrous. Any reasonable regional ISP, large or small, is going to go through that in a year. And if you enforce a 75% usage requirement to get more, then you've got something workable. I would define "usage" as "has a routing entry active on the net". Note that this does have an honesty component, as, for example, we have part-time networks connected via dial-up which are only routed when active. But trust me, we have issued what we asked for -- and that space IS being actively used by real, live, paying customers. The current arrogance of the NIC in asking for business plans and such is absolutely out of line and acts in restraint of trade. As such it has to go away, and the only way to get a monster which thinks it is God under control that works is to dice it up into smaller pieces, all of which KNOW they are not God. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
The request for a business plan was unreal. One of our customers was ready to demand the business plan for the last large provider who requested numbers. Has NET-99 been frustrated with the process ? Yes. Have we screwed up our SWIP's at times ? Yes. Has the Nic lost requests we sent in for domain names etc ? Yes. Is the Nic busy ? We all know that answer. More routes, more numbers more eveything is going to cost us in the end. When it comes to these numbers its time more of us had a say in controlling our own future. I support an increase in funding for the Nic. I have told them this many times. I also support the recall of unused or under-used numbers. By default I feel we the provider are in the position of telling our future customers why we are going to limit the amount of ip numbers we issue. Yes, I know SWIP them and they get more. Try telling that to a company who has just signed up and paid big dollars to do a nationwide 56k frame deal. Try listening to a customer saying he or she can get more numbers from another provider. Bottom line: Lets try to fix this in a common sense fashion and a reasonable time. It is possible. Joseph Stroup
i'm not sure i agree with karl here.
But the new ISP has no basis to request these. However, the CURRENT NIC is declining requests for *256* Class "C" addresses!
That is, I was turned down for a Class "B" equivalent in a CIDR block. That is ludicrous. Any reasonable regional ISP, large or small, is going to go through that in a year. And if you enforce a 75% usage requirement to get more, then you've got something workable.
this begs a question (which karl will proceed to try to answer), which is: how does the NIC know whether a new ISP is worthy of even 256 class c's? in karl's terminology, how do they know whether an ISP is "reasonable"? sure, "anybody knows," but i don't think this is an area where we can just depend on the NIC to decide based on their personal knowledge of the ISP or the industry. it's not that i don't trust them -- i do, actually. but the potential for liability on the NIC's part if they don't have -- and follow! -- objective allocation standards is too high to ask them to bear. the 75% usage problem is rather hard to verify, as well. to get a negative answer you'd have to know during the host count that the network being examined was at that moment routable. or you'd have to depend on the ISP to do its own host counts, which opens up horrendous fraud potential.
I would define "usage" as "has a routing entry active on the net".
i've got 9 bits of address space in the core routing tables right now. if i tell the NIC i'm planning to enter the ISP business, does my existing route entitle me to a block of 16 more bits? if not, why not? here's the crux of the problem:
Note that this does have an honesty component, as, for example, we have part-time networks connected via dial-up which are only routed when active. But trust me, we have issued what we asked for -- and that space IS being actively used by real, live, paying customers.
i know that. and if anybody asked me whether Net99 or MCSnet needed 16 bits per allocation, i would say "hell yes!" since i know you guys (a) know the meaning of what you're asking for, and (b) will use it wisely and honestly. i cannot depend on either (a) or (b) for the average new ISP-wannabe who has sold their video rental store and wants to reinvest the proceeds in the Internet 'cuz they saw Al Gore on Tee Vee and though they don't know what the Internet is, they know they gotta have some, and isn't that book by Canter and Seigel just the greatest thing you ever saw? hell no, i won't go. along that path lies chaos. what we need is some kind of "ISP Council", with core membership determined by some combination of customer votes and hard dollars (which keeps out ignorant newcomers and large providers whose customers don't like them but don't hate them enough to switch) and secondary membership by invitation of a majority of the core members (thus making room for people like kent england or even your humble author). this council would meet electronically every so often to ratify allocations which are outside a conservative upper limit, and meet physically every so often to argue about whether the underlying allocation scheme or current limits need to be changed. nothing i've seen on com-priv or cix-members or nanog or rfc1466bis has yet addressed the fundamental problem of using government money (or allowing any government to affect the policy) for internet resource allocation. we, the users (and the greater "we", the providers) have to take this over. we've got the dollars at risk and we've got the customers to satisfy -- and those two things are the primary components of networking.
we, the users (and the greater "we", the providers) have to take this over. we've got the dollars at risk and we've got the customers to satisfy -- and those two things are the primary components of networking.
Is it just me, or does this argument ring the same as the debate over Social Security and taxes? Yes, we all agree that the situation, if it proceeds as it has, it horrible. Yet, could we as the internet community surpass the lethargicness of our government? I believe our true test of autonomy will be demonstrated at this juncture. Do we choose to develop a plan that will prevent the european ags's from being driven to their deaths? Or shall we slide along in oblivion, getting blocks of 8 bits 4 or five times a year? At MIDnet we've just gotten another 16 bit CIDR, and soon we will likely need another one. Would it not have made more sense to give us an 18 bit CIDR? Regardless, many methods of implementing a better situation come to mind, and Vixie's "World Council of Internet Eleet" is far from the worst. -- Alan Hannan (402) 472-0241 MIDnet Inc. ------------------------------\ fax (402) 472-0240 A Global Internet Company " All perception of truth is \_________________________ the detection of an analogy " -- Henry David Thoreau \____________________
But the new ISP has no basis to request these. However, the CURRENT NIC is declining requests for *256* Class "C" addresses!
That is, I was turned down for a Class "B" equivalent in a CIDR block. That is ludicrous. Any reasonable regional ISP, large or small, is going to go through that in a year. And if you enforce a 75% usage requirement to get more, then you've got something workable.
this begs a question (which karl will proceed to try to answer), which is: how does the NIC know whether a new ISP is worthy of even 256 class c's? in karl's terminology, how do they know whether an ISP is "reasonable"?
Does it matter for the first block? I believe not. Or, if it does, then go ahead and say "ok, the first block for *those who don't have a link up and running at the time of request* is 32 Class "C"s", and the rest leave alone. Now, if the ISP wannabe goes nowhere, 32 Class "C"s isn't much. If you want to require an applicant to have a DBA/Corporate name, and prove it, that's fine too. But beyond that you're getting into restraint of trade. Its not any of the NIC's business what my business plan looks like!
but the potential for liability on the NIC's part if they don't have -- and follow! -- objective allocation standards is too high to ask them to bear.
But they aren't doing that NOW! Tell me that Sprint has a business plan on file with the NIC. Or Alternet, or PSI. Yeah, right. But others have reported to me that they have been asked for them, and I was a personal witness to one of those requests. If Sprint, PSI, and Alternet have them on file with the NIC I want to FOIA those documents! If not, then I allege that as of *right now* the NIC isn't playing on a level field, isn't being honest with its constituents, and is abusing its power. Power that badly needs to be removed and decentralized.
the 75% usage problem is rather hard to verify, as well. to get a negative answer you'd have to know during the host count that the network being examined was at that moment routable. or you'd have to depend on the ISP to do its own host counts, which opens up horrendous fraud potential.
Yep. It certainly does. But what other choice do you really have? The NIC is an arrogant organization now which thinks it is "God of the Addresses". Truth be known, its not, and never will be. It runs no route servers which could refuse to announce you. It doesn't even have real authority to delegate anything, other than the consent of those who are accepting addresses from it. A number of folks, myself included, are leaning towards revoking that consent or modifying it in drastic ways.
Note that this does have an honesty component, as, for example, we have part-time networks connected via dial-up which are only routed when active. But trust me, we have issued what we asked for -- and that space IS being actively used by real, live, paying customers.
i know that. and if anybody asked me whether Net99 or MCSnet needed 16 bits per allocation, i would say "hell yes!" since i know you guys (a) know the meaning of what you're asking for, and (b) will use it wisely and honestly. i cannot depend on either (a) or (b) for the average new ISP-wannabe who has sold their video rental store and wants to reinvest the proceeds in the Internet 'cuz they saw Al Gore on Tee Vee and though they don't know what the Internet is, they know they gotta have some, and isn't that book by Canter and Seigel just the greatest thing you ever saw?
Well, perhaps. But again, the newbie ISP problem can easily be solved by requiring anyone wanting that big allocation to have an *active* and *pingable* network core running at a reasonable bandwidth (SLIP need not apply :-). That solves the zero-eth order problem, which is the newbie coming in and requesting huge chunks of space they won't use. The other part of this would be to require a copy of the corporate annual report to be filed annually, and if you don't, any non-delegated space gets released. My *personal* issue with this is that when a firm I used to work for folded, the 32 Class "B"s they had allocated (with good reason I might add), were released back to the NIC -- by me personally. Less than two months later I asked for 256 Class "C"s and a "B" for internal network backbone use (two of the "B"s total) and was told to go to hell. The next time I get responsibility for 32 Class "B"s someone doesn't really need I'm going to SWIP the damn things to MCSNet and allocate my "C"s out of those -- that way I won't run out for a while, or deal with people telling me that my use isn't legitimate. The NIC can consider themself on notice for going out of their way to piss off someone who has, in the past, tried to HELP THEM.
hell no, i won't go. along that path lies chaos. what we need is some kind of "ISP Council", with core membership determined by some combination of customer votes and hard dollars (which keeps out ignorant newcomers and large providers whose customers don't like them but don't hate them enough to switch)
That might be ok; I don't know. The problem is that the council needs to be absolutely, 100% above reproach; it WILL get sued if anything that smells ugly comes out of it -- and eventually it will. This is why I like the non-profit, funded-by-users Internic idea; if you delegate in sufficiently large chunks, and charge a fee, then the fee is reasonably small (ie: $1k/year) which avoids restraint of trade problems AND if the Internic-to-be doesn't suit the members who are paying it, they stop paying (and set up something else).
nothing i've seen on com-priv or cix-members or nanog or rfc1466bis has yet addressed the fundamental problem of using government money (or allowing any government to affect the policy) for internet resource allocation. we, the users (and the greater "we", the providers) have to take this over. we've got the dollars at risk and we've got the customers to satisfy -- and those two things are the primary components of networking.
See above. A distributed allocation system with a reasonable fee structure at the top level going to a non-profit just might work. But *reasonable* means covering *only* the costs involved, and the key is to make the granularity sufficiently coarse that the fees are modest. This, by the way, is where the CIX failed. By refusing to tie costs to services rendered to real people it became a place where only "real" ISPs could play. This in and of itself wasn't a problem..... but then the *really big* ISPs then shut out the larger constituency and refused to listen to their voices -- even when expressed in the form of a membership vote at the annual meeting. How many of those disenfranchised ISPs renewed this Jan 1st? And of those who did, how many felt and continue to feel like hostages and victims of extortion? The Internic issues are TOO important to let anything like the CIX happen to them. If they do the end result *will* be government involvement, and not in a fashion that you, or I, would approve of. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
On Sun, 19 Mar 1995, Karl Denninger wrote:
The NIC can consider themself on notice for going out of their way to piss off someone who has, in the past, tried to HELP THEM.
I too have tried to responsibly recycle just class-C networks and found that their position and attitude is worthless to work with. I try to recycle any- thing where I can, but this is getting to be too much like typing paper that costs twice as much recycled as compared to non-recycled stuff. In other words, I'm willing to go to as much effort to recycle addresses as get new ones, but push me much further and I'm not going to bother. My last experience was with two class-C nets -- something most folks probably wouldn't even bother with -- and ended up wasting a bunch of time with e-mail exchanges, researching, etc. So, preach it Karl... Ed Morin Northwest Nexus, Inc.
I have an idea. Why doesn't someone (a big network provider or some other entity which has technically competent people) offer the NIC $10,000 for a huge block of CIDR addresses which it will then resell (with prompt assignment and service etc.) to providers which need addresses, presumably at about the price they paid for them plus admin costs plus 20% or so. -george
I have an idea. Why doesn't someone (a big network provider or some other entity which has technically competent people) offer the NIC $10,000 for a huge block of CIDR addresses which it will then resell (with prompt assignment and service etc.) to providers which need addresses, presumably at about the price they paid for them plus admin costs plus 20% or so.
I know of several companies that would be willing to spend over $10,000 for a class B network. Money is not the issue. Your idea of a (more efficient) distribution center makes sense, however, this should be achieved through group concensus and appointed councils, not by who wants to make money. -- Alan Hannan (402) 472-0241 MIDnet Inc. ------------------------------\ fax (402) 472-0240 A Global Internet Company " All perception of truth is \_________________________ the detection of an analogy " -- Henry David Thoreau \____________________
Group consensus several months ago (in another discussion group, com-priv) was that it would be a good idea for CIX and/or other interested groups to send people to work at InterNIC just to get the workload down and response time up, which would also facilitate dialog with InterNIC about their CIDR block policy (hopefully). This appears to have gone nowhere... People have been saying this is a horrible situation for six months. That's too long. I think that any solution which won't in and of itself create worse long term problems is looking attractive now. I don't see any such problems with another agency acting as a CIDR vendor for a limited set of addresses, while we look for long term solutions. -george william herbert gherbert@crl.com
From: Alan Hannan <alan@mid.net> Subject: Re: Internic address allocation policy I know of several companies that would be willing to spend over $10,000 for a class B network. Money is not the issue. Your idea of a (more efficient) distribution center makes sense, however, this should be achieved through group concensus and appointed councils, not by who wants to make money. If T.J. Rogers was dead, he'd be spinning in his grave right now. If you want something done efficiently, you want it run by people who are damn well hungry for money, not by appointed councils.
If T.J. Rogers was dead, he'd be spinning in his grave right now.
If you want something done efficiently, you want it run by people who are damn well hungry for money, not by appointed councils.
In most economic systems I would agree with you. I do heartily believe in supply and demand. However, I believe that rationing and hoarding cause exceptions to this law of economics. My capitalism knows few bounds, but when it comes to IP numbers, the only way I want money to be directly tied to them is if I'm the one with the control ;). I very much like Vixie's idea of a committee to allow or disallow appropriation of IP numbers. I do not like the idea of selling the numbers to people. MIDnet and other ISPs should earn their IP blocks by describing what they intend to do with them, and how they intend to implement them efficiently. This should not be a money issue, as the supply is (relatively) small, and the ownership starts upon assignment. -- Alan Hannan (402) 472-0241 MIDnet Inc. ------------------------------\ fax (402) 472-0240 A Global Internet Company " All perception of truth is \_________________________ the detection of an analogy " -- Henry David Thoreau \____________________
I believe that people are indeed already describing what they intend to do with them and how they will impliment things efficiently. The problem appears to be that that isn't enough. Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient chunks and leaves the vast majority of class C space open for whatever long term solution we come up with. It's not like we're going to run out of C's soon, far from it last time I looked. That would be 1/32 of the class C space if I remember where the top end is correctly. Class C space is nearly empty right now. We can afford to use that in a quick fix. -george william herbert gherbert@crl.com
First, may I ask that when you reply to a message from the nanog mailing list, you edit the headers so that they say "To: nanog@merit.edu" and have no CC? Right now there is a strong penalty for anyone who adds to a thread, since we will be on the CC list forever (getting two copies) even when it moves to a different topic. We are all on the nanog list, no need to CC us. Second, I've seen Karl and now Alan misuse a term. I'll pick on Alan since his message is right in front of me, but the complaint is general (sorry Alan!):
Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient
That's 16M addresses, not 64K addresses. We should not equivocate "addresses" and "Class C networks". 210.*.*.* has 2^24 (minus subnet zero and broadcast lossage) addresses -- 16M. 210.*.*.* has 2^16 "Class C networks" -- 64K. We must not assume that every customer will get a Class C -- many will get just a subnet since they will only have a handful of hosts. I know of several providers who are chopping things up on nybble boundaries (16 hosts/net, or actually 14 with the subnet zero and broadcast taken out).
Paul A Vixie
Second, I've seen Karl and now Alan misuse a term. I'll pick on Alan since his message is right in front of me, but the complaint is general (sorry Alan!):
Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient
That's 16M addresses, not 64K addresses. We should not equivocate "addresses" and "Class C networks". 210.*.*.* has 2^24 (minus subnet zero and broadcast lossage) addresses -- 16M. 210.*.*.* has 2^16 "Class C networks" -- 64K. We must not assume that every customer will get a Class C -- many will get just a subnet since they will only have a handful of hosts. I know of several providers who are chopping things up on nybble boundaries (16 hosts/net, or actually 14 with the subnet zero and broadcast taken out).
Yep, I did this when I worked for SCCSI in Houston. We also had to go through the much-argued address allocation summary, or *PLAN*, with Sesquinet. I don't see what everyone is fussing about. I made myself nuts for a while trying to guess *exactly* what my network would look like in 2 years, then 5... and then I just gave up. They said to estimate, so I estimated. And I got the address space. Granted, I never had to deal directly with the NIC but Sesquinet was pretty strict and I still got space. They also implemented the IETF recommended formula for determining usage statistics. There were a certain number of hosts on your network which had to be "pingable" and another percentage that had to be listed in DNS. Now, our problem was that we gave out class C addresses subnetted with 255.255.255.252 (yes we wasted half the address space!) for SLIP/PPP accounts. So we were under our percentage of "pingable" hosts. Once I explained why, though, we were ok'd for more addresses. It just seems to me that everyone is taking all of this way out of bounds. The internic is strict and sometimes contradicts itself, I agree. but all they ask for is a plan, or rather and *estimate*. Karl, you say you can't possibly guess those numbers? How do you do business then? You've got to be able to estimate business and market growth (not bashing you, just don't understand). -brett
First, may I ask that when you reply to a message from the nanog mailing list, you edit the headers so that they say "To: nanog@merit.edu" and have no CC? Right now there is a strong penalty for anyone who adds to a thread, since we will be on the CC list forever (getting two copies) even when it moves to a different topic. We are all on the nanog list, no need to CC us.
Ok :-)
Second, I've seen Karl and now Alan misuse a term. I'll pick on Alan since his message is right in front of me, but the complaint is general (sorry Alan!):
Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient
That's 16M addresses, not 64K addresses. We should not equivocate "addresses" and "Class C networks". 210.*.*.* has 2^24 (minus subnet zero and broadcast lossage) addresses -- 16M. 210.*.*.* has 2^16 "Class C networks" -- 64K. We must not assume that every customer will get a Class C -- many will get just a subnet since they will only have a handful of hosts. I know of several providers who are chopping things up on nybble boundaries (16 hosts/net, or actually 14 with the subnet zero and broadcast taken out).
Not me! I consider a "Class B equivalent" to be 256 NETWORKS, by the common use of the term, but 65K *addresses*. For many of our account classifications we chop up network numbers to significantly smaller pieces. For some of them we chop them up to the *address* level, that is, ONE ADDRESS (1.2.3.4). For others we chop it up to the four-address level (mask 255.255.255.252), yielding three usable hosts if you can use "zero" (two if not) and for others we hand out larger blocks, including full Class "C"s. Right now, for political reasons, anything that a customer can take away from us is a class "C", primarily because not all providers are CIDR capable and the SWIP process is not really good for sub-C blocks, and until the backbone folks start insisting on this (we do for resale connections) this is the only way to insure that a customer doesn't get screwed somewhere down the line. A customer with a single SLIP connection gets one *address*; that is, 1.2.3.4, and that's it. Our backbone routes that internally. Yes, that is hell on the routers. However, it is also efficient on address space. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
Date: Mon, 20 Mar 1995 11:21:12 -0600 (CST) From: karl@mcs.com (Karl Denninger) To: nanog@merit.edu
Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient
That's 16M addresses, not 64K addresses. We should not equivocate "addre sses" and "Class C networks". 210.*.*.* has 2^24 (minus subnet zero and broadc ast lossage) addresses -- 16M. 210.*.*.* has 2^16 "Class C networks" -- 64K. We must not assume that every customer will get a Class C -- many will get j ust a subnet since they will only have a handful of hosts. I know of several providers who are chopping things up on nybble boundaries (16 hosts/net, or actually 14 with the subnet zero and broadcast taken out).
Not me!
I consider a "Class B equivalent" to be 256 NETWORKS, by the common use of the term, but 65K *addresses*.
1 Class-B-sized prefix = 256 Class-C-sized prefixes. A "network" could be a Class A, B, C, etc., network, or a CIDR network, etc.
Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs. net
Mr. Vixie wrote:
Second, I've seen Karl and now Alan misuse a term. I'll pick on Alan since his message is right in front of me, but the complaint is general (sorry Alan!):
Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient
Not to add insult, but that message was +_to_+ me, not +_from_+ me... ;) I do fully understand the fact that a class C network is nothing other than an arbitrarily sized network. This is part of the general problem causing reaction by the Internic, I believe. They have perhaps reacted too strongly fearing that all people who come with plans have _not_ considered subnetting, thereby penalizing the whole, when without argument there have been people who have adequately planned efficiency, but been told they had not. Whenever I get new customers, I explain to them the joys of subnetting. Often they say it would be easier to get a class B network. I laugh and say it would be easier to get a class A but it's not going to happen. The only question here is trading convenience for efficiency, at least to the person applying. And people will rarely choose efficiency voluntarily, hence the Internic/applicant quarrel. So, we run into two seperate discussions: 1) How do we ration IP numbers? 2) How do we aggregate IP numbers. This discussion has taken place several times, but I offer my two cents: This proposal embodies pieces of both George Herbert and Vixie's ideas. We split off, oh say, 3 Class A sized CIDR blocks. Split these into, oh, say 24 Aggregate blocks, each of which having the 13 bits of subnet mask, and 19 bits of network space. Then we assign these 24 blocks to "responsible" ISPs. These ISPs would then be tasked with efficiently managing these networks until, oh, say the year 2000. By then IPng will be out, right? ;) Perhaps we'd vary the size of the blocks, perhaps not, regardless, this delegates out the workload and should stop the finger pointing, or at least make it local. The kicker is that I would like to see some sort of arbitration group heading up the "Fairness" of the allocation. Whether this is the Internic, or Vixie's World Internet Address Allocation Committee, someone in place to see that DUMBOnet isn't mismanaging their IPs, and that they aren't charging too much for the administration costs. Comments? -- Alan Hannan (402) 472-0241 MIDnet Inc. ------------------------------\ fax (402) 472-0240 A Global Internet Company " All perception of truth is \_________________________ the detection of an analogy " -- Henry David Thoreau \____________________
Paul writes:
Second, I've seen Karl and now Alan misuse a term. I'll pick on Alan since his message is right in front of me, but the complaint is general (sorry Alan!):
That was me, actually.
Taking a relatively small chunk of the remaining address space (say, 210.*.*.*) gives us 64k addresses to hand out in convenient
That's 16M addresses, not 64K addresses. We should not equivocate "addresses" and "Class C networks". 210.*.*.* has 2^24 (minus subnet zero and broadcast lossage) addresses -- 16M. 210.*.*.* has 2^16 "Class C networks" -- 64K. We must not assume that every customer will get a Class C -- many will get just a subnet since they will only have a handful of hosts. I know of several providers who are chopping things up on nybble boundaries (16 hosts/net, or actually 14 with the subnet zero and broadcast taken out).
I slipped. It's 64k class C networks. I know better, but yesterday was a long day. If all the router vendors supported nybble-sized routing, things would be a lot easier for providers. If there was an easy named db syntax to fix in-addr mapping syntax for nybble-sized routing, things would be a lot easier for providers. Paul can perhaps fix one of these issues (in his copious spare time? 8-), the other one is a more general problem. -george william herbert gherbert@crl.com
Paul, I agree with you. But, we have a problem and my point was there are enough of us out here to solve it. The one thing that I enjoyed about the NANOG meeting was meeting the people. I was very impressed with the technical level these people work at. As a group we can and will solve this. Joseph Stroup
Uh, aside, as you get community consensus, you obviously need to also get consensus with the *future* community, as people seem to fault ARPA for not having considered 1995's Internet community, and grabbed authority where they did not have any, when they published the IP spec in September 1981 as part of their research agenda, and from where the IP allocation grew.
Han, We are NOT trying to hassle the Nic. And yes they do work under the guidelines. What we are saying is there is a real problem and by talking about it here, enough people have come forward to verify it. I will be more than happy to act as the contact with all parties involved and get the ball rolling. As I said before, my goal is to make this an easier process. I have telephones and they need numbers. I am not alone. This has been a good discussion. Time to move forward with some action.
On Sun, 19 Mar 1995, Karl Denninger wrote:
My request was for 256 Class "C"s. We have consumed almost exactly that in a year.
I did the same. With a little drawing and a bit of written proof, I had no problem in getting the 199.72.0.0 block for Interpath. Interpath has not *QUITE* used up the block, but pretty darn close, as Michael and others can attest... I certainly believe that if you are not willing to produce proof [of some sort], you should not just be given the resources.... Heck, why not a policy of "proof or pay". If you are not willing to show that you are seriously planning to use them, *THEN* you pay for them? Sorry to disagree with you on this one, Karl... 8-) -abc Alan B. Clegg Information Systems Manager American Research Group
On Sun, 19 Mar 1995, Karl Denninger wrote:
My request was for 256 Class "C"s. We have consumed almost exactly that in a year.
I did the same. With a little drawing and a bit of written proof, I had no problem in getting the 199.72.0.0 block for Interpath. Interpath has not *QUITE* used up the block, but pretty darn close, as Michael and others can attest...
I certainly believe that if you are not willing to produce proof [of some sort], you should not just be given the resources.... Heck, why not a policy of "proof or pay". If you are not willing to show that you are seriously planning to use them, *THEN* you pay for them?
Sorry to disagree with you on this one, Karl... 8-) -abc
Alan B. Clegg Information Systems Manager American Research Group
Heh, I can lie my way through anything. I just refuse to do so. Can I predict a year out where we will have POPs, what kind of customers will be behind those POPs, or where they will be situated and how we will route their networks for them? No damn way. No ISP in the business can possibly do that and be telling the truth. I can tell you what I have on my desk in the form of orders *right now*, but tomorrow is another day, and when you're growing 20% a month it is absolutely impossible to predict anything other than a gross guess at what your actual requirements are. As it turns out, I'm a good guesser and historical evidence points this out -- both now and a year ago. But the level of detail that the NIC asked me for was *impossible* to provide unless I just wanted to lie and generally play with people's heads. I don't play that game. I've been installing networks for people for damn near 10 years, and on the net for eight of those. I've worked with the NIC to get things done for those firms I worked for, and my own, for most of that time. The folks there know damn well who I am and what I build, and by now it should be irrelavent to request my business plan -- unless they want it for some other, not-so-kosher reason! -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
On Mon, 20 Mar 1995, Karl Denninger wrote:
Heh, I can lie my way through anything. I just refuse to do so.
Yup.. I guess that when I drew it out in detail with only two mistakes [POPs ended up a different location than what was planned] I was lying. Nope, I did not lie, and neither would you if you were to think a little. A business plan is just that, a PLAN, and the NIC is asking for a PLAN, not a full view of the future. If you are wrong, you are wrong. If you submit what you PLAN, that is what they are asking for.
Can I predict a year out where we will have POPs, what kind of customers will be behind those POPs, or where they will be situated and how we will route their networks for them?
If not, you don't have an idea of what your business is going to be doing. As I said above, you may plan wrong, but you sure as HELL should have a plan that at least goes a year into the future.
No damn way. No ISP in the business can possibly do that and be telling the truth.
I don't appreciate being called a liar, Karl. They asked for a PLAN, I supplied a PLAN. -abc Alan B. Clegg Information Systems Manager American Research Group
On Mon, 20 Mar 1995, Karl Denninger wrote:
Heh, I can lie my way through anything. I just refuse to do so.
Yup.. I guess that when I drew it out in detail with only two mistakes [POPs ended up a different location than what was planned] I was lying.
Nope, I did not lie, and neither would you if you were to think a little. A business plan is just that, a PLAN, and the NIC is asking for a PLAN, not a full view of the future. If you are wrong, you are wrong. If you submit what you PLAN, that is what they are asking for.
MCSNet's business plan is CONFIDENTIAL. When the NIC is willing to sign a $10M indemnity guarantee of non-release, standard business confidentiality agreements, and provide a list of EXACTLY who is viewing it, and why, then I might release one. Until then it is none of their damn business, and that is and was my exact response to the request.
Can I predict a year out where we will have POPs, what kind of customers will be behind those POPs, or where they will be situated and how we will route their networks for them?
If not, you don't have an idea of what your business is going to be doing. As I said above, you may plan wrong, but you sure as HELL should have a plan that at least goes a year into the future.
I have a business plan that goes *five* years out into the future, and nobody -- but nobody -- is ever going to see it without DAMN good justification.
No damn way. No ISP in the business can possibly do that and be telling the truth.
I don't appreciate being called a liar, Karl.
They asked for a PLAN, I supplied a PLAN.
-abc
Alan B. Clegg Information Systems Manager American Research Group
Then you're a fool, almost as bad as the one who is announcing OSPF routes at the MAE. Anyone who gives business plans out to people without iron-clad confidentiality guarantees deserves to get screwed. If your business plan is "real", in that it contains the major elements of any professional business plan (ie: your marketing strategies, your intended customer base, your growth projections and geographic interests, etc), then that information is likely to be some of the most valuable that you develop and posess. In fact, I would argue that your business plan is more important than ANY amount of technology you develop. If you're willing to hand that out to people, or get extorted into giving it out, that's your choice, but its certainly not mine, and certainly not that of anyone who, IMHO, has an operational brain in their head as regards business and how it functions. Since you feel that it is of so little value that the NIC could have a copy without any signed guarantee of non-release, how about you post a copy here for our edification so I know where to set my next few POPs -- right on top of yours. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
I am told by someone with direct knowledge that all they want is some estimate of numbers of pops and methods of connecting them. IE some data about proposed network topology and nothing more and that they pledge to keep this confidential. Karl defined what was wanted in rather broader terms unless I misunderstood his message. Why can't some general estimate at this fairly high and abstract level with geographic locations of POPS not precisely pin pointed (pick the radius that suits you - 10 miles - 5- or 50 miles) be given. Several people directly involved in the process are telling me that such general and fairly abstract data is quite sufficient and that they absolutely do NOT want anything to do with how the ISP defines it future market. Karl: is it your point that you cannot even give general data without giving away more than you want about market you are pursuing? ******************************************************************** Gordon Cook, Editor & Publisher Subscript.: Individ-ascii $85 The COOK Report on Internet -> NREN Non Profit. $150 431 Greenway Ave, Ewing, NJ 08618 Small Corp & Gov't $200 (609) 882-2572 Corporate $350 Internet: cook@cookreport.com Corporate. Site Lic $650 ******************************************************************** On Mon, 20 Mar 1995, Karl Denninger wrote:
On Mon, 20 Mar 1995, Karl Denninger wrote:
Heh, I can lie my way through anything. I just refuse to do so.
Yup.. I guess that when I drew it out in detail with only two mistakes [POPs ended up a different location than what was planned] I was lying.
Nope, I did not lie, and neither would you if you were to think a little. A business plan is just that, a PLAN, and the NIC is asking for a PLAN, not a full view of the future. If you are wrong, you are wrong. If you submit what you PLAN, that is what they are asking for.
MCSNet's business plan is CONFIDENTIAL. When the NIC is willing to sign a $10M indemnity guarantee of non-release, standard business confidentiality agreements, and provide a list of EXACTLY who is viewing it, and why, then I might release one.
Until then it is none of their damn business, and that is and was my exact response to the request.
Can I predict a year out where we will have POPs, what kind of customers will be behind those POPs, or where they will be situated and how we will route their networks for them?
If not, you don't have an idea of what your business is going to be doing. As I said above, you may plan wrong, but you sure as HELL should have a plan that at least goes a year into the future.
I have a business plan that goes *five* years out into the future, and nobody -- but nobody -- is ever going to see it without DAMN good justification.
No damn way. No ISP in the business can possibly do that and be telling the truth.
I don't appreciate being called a liar, Karl.
They asked for a PLAN, I supplied a PLAN.
-abc
Alan B. Clegg Information Systems Manager American Research Group
Then you're a fool, almost as bad as the one who is announcing OSPF routes at the MAE.
Anyone who gives business plans out to people without iron-clad confidentiality guarantees deserves to get screwed. If your business plan is "real", in that it contains the major elements of any professional business plan (ie: your marketing strategies, your intended customer base, your growth projections and geographic interests, etc), then that information is likely to be some of the most valuable that you develop and posess. In fact, I would argue that your business plan is more important than ANY amount of technology you develop.
If you're willing to hand that out to people, or get extorted into giving it out, that's your choice, but its certainly not mine, and certainly not that of anyone who, IMHO, has an operational brain in their head as regards business and how it functions. Since you feel that it is of so little value that the NIC could have a copy without any signed guarantee of non-release, how about you post a copy here for our edification so I know where to set my next few POPs -- right on top of yours.
-- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
I am told by someone with direct knowledge that all they want is some estimate of numbers of pops and methods of connecting them. IE some data about proposed network topology and nothing more and that they pledge to keep this confidential.
Karl defined what was wanted in rather broader terms unless I misunderstood his message. Why can't some general estimate at this fairly high and abstract level with geographic locations of POPS not precisely pin pointed (pick the radius that suits you - 10 miles - 5- or 50 miles) be given. Several people directly involved in the process are telling me that such general and fairly abstract data is quite sufficient and that they absolutely do NOT want anything to do with how the ISP defines it future market.
Karl: is it your point that you cannot even give general data without giving away more than you want about market you are pursuing?
******************************************************************** Gordon Cook, Editor & Publisher Subscript.: Individ-ascii $85
No, we provided them with that general data and were told that wasn't enough information for them to make a decision. The general case is publically known and isn't something that is considered proprietary; the specific case is, and is not something we're about to disclose. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
What effect will the change of ownership at the NIC have? I for one would be wary of giving them information that I want them to keep private; they are not a government agency who is at least bound by some rules. Now that the NIC has been sold to ______ (I have no idea if I am allowed to say so I will put it in as a blank, but I am pretty sure this is all public knowledge) would agreements to keep things private really be secure with a new owner, etc? Doug
Karl wrote:
My request was for 256 Class "C"s. We have consumed almost exactly that in a year.
And Alan added:
I did the same. [...] Interpath has not *QUITE* used up the block, but pretty darn close
Can either of you provide estimates of how densely these blocks are populated? What percentage of the individual host addresses are currently used by functioning hosts (or router interfaces)? Perhaps you're both saying you re-assigned use of these address blocks quickly, rather than saying you populated them densely with actual hosts? One could imagine ISPs (not Karl or Alan, surely!) who might re-assign blocks of class C (octet-wide) networks when a single one would suffice. Or they might assign octet-wide networks when quartet-wide (nibble) networks would suffice; or even quartet-wide networks when triplet-wide networks would suffice. I believe ISPs who receive large blocks have a responsibility to make sure they (and their customers) use them efficiently. How efficiently, you ask? If you're not getting more than one U.S. dollar of value per host address per year, you're not using it efficiently in my book! -- Sean Shapira sds@jazzie.com +1 206 443 2028 <a href="http://www.jazzie.com/sds/">Sean's Home Page</a> Serving the Net since 1990.
Karl wrote:
My request was for 256 Class "C"s. We have consumed almost exactly that in a year.
And Alan added:
I did the same. [...] Interpath has not *QUITE* used up the block, but pretty darn close
Can either of you provide estimates of how densely these blocks are populated? What percentage of the individual host addresses are currently used by functioning hosts (or router interfaces)?
Of course -both- of these ISP's are running RWhois servers which accuratly reflect the delegations that they have made and which can be used to document to the Internic that they need more space. (beats SWIP entries all hollow IMHO) -- --bill
Karl wrote:
My request was for 256 Class "C"s. We have consumed almost exactly that in a year.
And Alan added:
I did the same. [...] Interpath has not *QUITE* used up the block, but pretty darn close
Can either of you provide estimates of how densely these blocks are populated? What percentage of the individual host addresses are currently used by functioning hosts (or router interfaces)?
In some cases, 99%. Really. We hand out single addresses and route them in many cases, and those addresses get used until they're gone -- all of them, except for all zeros and all ones. Some blocks are less densely populated. If I subnet to 26 bits then 1/2 of that Class "C" is wasted, which I can't control due to the technical requirements of the network involved.
Perhaps you're both saying you re-assigned use of these address blocks quickly, rather than saying you populated them densely with actual hosts?
Uh, no. We populated them as densely as reasonably possible given the requirements of our customers and our own internal engineering allows. We have spent a hell of a lot of money, and made a number of purchasing decisions, based on whether or not certain products could or could not do VLSMs and the like, and whether they could handle host routes. All in the name of not wasting address space.
One could imagine ISPs (not Karl or Alan, surely!) who might re-assign blocks of class C (octet-wide) networks when a single one would suffice. Or they might assign octet-wide networks when quartet-wide (nibble) networks would suffice; or even quartet-wide networks when triplet-wide networks would suffice.
Until there is a solution for in-addr.arpa quartet-wide or triplet-wide network assignments to individual customers who want to run their own DNS (and many do for VERY good reasons) is damn close to impossible. This is a technical problem which has not been solved in the DNS system for a long time, and it stinks, but its a reality of life right now. If and when it is fixed we'll adjust our allocation policies to match the new reality of the world. Until then we have to live with the reality as it sits. You want to know what would, more than anything else, help the address space issue? Solving the in-addr.arpa problem. Solve that one and I can reduce a number of my customer allocations, and so can all the other providers. That could, I would estimate, cut the growth rate of address requests at least in half.
I believe ISPs who receive large blocks have a responsibility to make sure they (and their customers) use them efficiently. How efficiently, you ask? If you're not getting more than one U.S. dollar of value per host address per year, you're not using it efficiently in my book! -- Sean Shapira sds@jazzie.com +1 206 443 2028 <a href="http://www.jazzie.com/sds/">Sean's Home Page</a> Serving the Net since 1990.
We're getting significantly more than one US dollar of revenue per host address per year from our allocations. In much of our address space we're exceeding that metric by more than two orders of magnitude. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
with ciscos ;you can use subnets all ones and all zeroes. So in fact you can use 2 subnmets in one class C with the 'bad' mask 255.255.128.0 . Maybe the other vendors should do that too, it is a question of customer pressure. Explain why you need it and why it is possible without creating chaos, and the vendor will certainly do it (Cisco did for al lones, e.g.) Mike On Mon, 20 Mar 1995, Karl Denninger wrote:
Karl wrote:
My request was for 256 Class "C"s. We have consumed almost exactly that in a year.
And Alan added:
I did the same. [...] Interpath has not *QUITE* used up the block, but pretty darn close
Can either of you provide estimates of how densely these blocks are populated? What percentage of the individual host addresses are currently used by functioning hosts (or router interfaces)?
In some cases, 99%. Really. We hand out single addresses and route them in many cases, and those addresses get used until they're gone -- all of them, except for all zeros and all ones.
Some blocks are less densely populated. If I subnet to 26 bits then 1/2 of that Class "C" is wasted, which I can't control due to the technical requirements of the network involved.
Perhaps you're both saying you re-assigned use of these address blocks quickly, rather than saying you populated them densely with actual hosts?
Uh, no. We populated them as densely as reasonably possible given the requirements of our customers and our own internal engineering allows. We have spent a hell of a lot of money, and made a number of purchasing decisions, based on whether or not certain products could or could not do VLSMs and the like, and whether they could handle host routes. All in the name of not wasting address space.
One could imagine ISPs (not Karl or Alan, surely!) who might re-assign blocks of class C (octet-wide) networks when a single one would suffice. Or they might assign octet-wide networks when quartet-wide (nibble) networks would suffice; or even quartet-wide networks when triplet-wide networks would suffice.
Until there is a solution for in-addr.arpa quartet-wide or triplet-wide network assignments to individual customers who want to run their own DNS (and many do for VERY good reasons) is damn close to impossible. This is a technical problem which has not been solved in the DNS system for a long time, and it stinks, but its a reality of life right now.
If and when it is fixed we'll adjust our allocation policies to match the new reality of the world. Until then we have to live with the reality as it sits.
You want to know what would, more than anything else, help the address space issue? Solving the in-addr.arpa problem. Solve that one and I can reduce a number of my customer allocations, and so can all the other providers. That could, I would estimate, cut the growth rate of address requests at least in half.
I believe ISPs who receive large blocks have a responsibility to make sure they (and their customers) use them efficiently. How efficiently, you ask? If you're not getting more than one U.S. dollar of value per host address per year, you're not using it efficiently in my book! -- Sean Shapira sds@jazzie.com +1 206 443 2028 <a href="http://www.jazzie.com/sds/">Sean's Home Page</a> Serving the Net since 1990.
We're getting significantly more than one US dollar of revenue per host address per year from our allocations. In much of our address space we're exceeding that metric by more than two orders of magnitude.
-- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity Modem: [+1 312 248-0900] | (shell, PPP, SLIP, leased) in Chicagoland Voice: [+1 312 248-8649] | 6 POPs online through Chicago, all 28.8 Fax: [+1 312 248-9865] | Email to "info@mcs.net" for more information ISDN: Surf at Smokin' Speed | WWW: http://www.mcs.net, gopher: gopher.mcs.net
--------------------------------------------------------------------------------- Michael F. Nittmann nittmann@wis.com Network Architect B3 Corporation (715) 387 1700 xt. 158 ---------------------------------------------------------------------------------
From: Mike Nittmann <nittmann@wis.com> Subject: Re: Efficient (Dense) Use of Address Blocks with ciscos, you can use subnets all ones and all zeroes. We recently changed (a year or so ago) to allow use of the all ones. We still make you turn a knob (ip subnet-zero) before allowing you to use subnet zero since some hosts and some vendor's implementations of maskless routing protocols barf when you use subnet zero. The current router requirements now specifies that the all zeros and all ones subnets are totally legal.
If you fail to agree, you don't get a number. Throughout the FCC letters there is a constant mention of "this scarce DNIC resource may necessitate reassignment of codes not implemented within a six-month period.
Now if the FCC has been doing this I am sure we could work out something with the Nic.
Joseph Stroup
Of course folks in China recognise the authority of the FCC in this matter. :) Or even the Internic... -- --bill
If you fail to agree, you don't get a number. Throughout the FCC letters there is a constant mention of "this scarce DNIC resource may necessitate reassignment of codes not implemented within a six-month period.
Now if the FCC has been doing this I am sure we could work out something with the Nic.
Joseph Stroup
Of course folks in China recognise the authority of the FCC in this matter. :) Or even the Internic...
Actually you can bet China does comply with international treaty and local regulation when attaching up to the International X.25 Network. And you can also bet they they get their addresses assigned by the Nic and in cooperation with with Asia IP coordinators. One could hardly call them connected if they couldn't complete X.25 calls, or route IP to everyone else in the world. The point here is that, the Internic is being counter productive, with the expressed goals of CIDR. They have not been assigning large enough blocks to ISPs, which defeats the purpose of CIDR. There are only two reasons I can see for this: The Internic doesn't have a clue as to what the hell they are doing. OR The are intentionally abusing a trust granted them by the Internet community. If such policies continue, that trust WILL be revoked, either through action internal (IANA, IETF, and others) or external, litigation. Either way they are asking for trouble, because when it comes right down to it, if I cannot, through reasonable efforts on my part, be assigned the IP address I need for my business, there is going to be a stink. And you can bet the 500+ other small ISPs are going to take this active threat to their livelyhood sitting down. -- ---------------------------------------------------------------------------- | Jeremy Porter (512)-339-6094 Freeside Communications, Inc. info@fc.net | | jerry@fc.net (512)-339-4466 (data) P.O. Box 530264 Austin, TX 78753 | ----------------------------------------------------------------------------
The point of this was the "scarce resource" part. joseph On Sat, 18 Mar 1995, Bill Manning wrote:
If you fail to agree, you don't get a number. Throughout the FCC letters there is a constant mention of "this scarce DNIC resource may necessitate reassignment of codes not implemented within a six-month period.
Now if the FCC has been doing this I am sure we could work out something with the Nic.
Joseph Stroup
Of course folks in China recognise the authority of the FCC in this matter. :) Or even the Internic...
-- --bill
participants (20)
-
Alan B. Clegg -
Alan Hannan -
ATM_Feel_the_Power -
bmanning＠ISI.EDU -
Brett Watson -
Doug Humphrey -
Ed Morin -
George Herbert -
Gordon Cook -
hwb＠upeksa.sdsc.edu -
Jeremy Porter -
karl＠mcs.com -
Michael S. Ramsey -
Mike Nittmann -
Paul A Vixie -
Paul Lustgraaf -
Paul Traina -
Paul Traina -
sds＠jazzie.com -
Steven J. Richardson