- NANOG - lists.nanog.org

Allocation of IP Addresses
by Jim Browning 14 Mar '96

14 Mar '96

I may be opening a can of worms, and if so, I believe it is one which needs to be opened. If this topic has been beaten to death in the past, then I apologize, however as it is a rapidly evolving topic, it warrants repeated discussion and evaluation. My fundamental questions are: 1. Is InterNIC consistently applying objective criteria in its evaluation of requests for the allocation of IP address blocks? 2. If so, what are the criteria? The INTERNIC IP ALLOCATION GUIDELINES FOR INTERNET SERVICE PROVIDERS states that "allocation is based on the ISP's 3 - 6 month requirement and other information the InterNIC deems necessary". There is no detail provided (in any document I have found) of what other information InterNIC deems to be necessary. I find an apparent conflict between established policy and its actual implementation on a day-to-day basis. CIDR dictates that addresses should be aggregated into the largest blocks possible, and that the publishing of extraneous routes be eliminated. In keeping with this, and because of often discussed operational considerations, the minimum size of blocks routed at the NAPs is growing larger and larger. To ease participation at the national level, you must ensure to the fullest extent possible that your address space is routable as a single block. In order to accomplish this, you must obtain either: A. a single allocation capable of supporting planned growth, or B. incremental allocations of *contiguous* blocks InterNIC's current CIDR allocation practice does not support either of these options. Due to the shortage of *available* IP addresses (there are of course millions of allocated but unused addresses floating around), InterNIC is using a "slow start" approach which provides incremental increases in total address space, with no guarantee that future increments will be contiguous. This means that the only way to maintain efficient routing is to engage in repeated renumbering of customer addresses to consolidate into increasingly larger blocks. How many times is it reasonable to ask a customer to renumber? Once is certainly reasonable. Twice is questionable. More than that and I would suspect the customer would renumber all right, but as part of shifting to a different ISP. The day to day implementation of policy by the InterNIC has increasingly critical impact on our industry, to the point of controlling who has the opportunity to succeed and who does not. IMHO, it is imperative that: 1. this function be performed in an understandable manner, 2. objective criteria be consistently applied 3. the criteria in use be publicly available, and 4. there be defined mechanisms for the 'appeal' of decisions made in the processing of allocation requests. Recent experience and observation leads me to conclude that these imperatives are perhaps not being met. Am I all wet???? -- Jim Browning <jfbb(a)ATMnet.net> 619/812-2860 Fax 619/812-2867

12 19

Re: Allocation of IP Addresses
by David C. Kovar 14 Mar '96

14 Mar '96

At 01:15 PM 3/14/96 -0500, Paul Ferguson wrote: >I've moved my response and this thread to CIDRD, since it appears to be >more appropriate in this venue. > >- paul Where is the CIDRD mailing list? Thanks! -David

1 0

Re: Allocation of IP Addresses
by owen＠DeLong.SJ.CA.US 14 Mar '96

14 Mar '96

> Hi, > > >Here's an idea. Let new ISP's reserve large blocks (say /16's) in 65/8, > >66/8, .... but don't let them actually use these addresses on the global > >Internet. Then, the ISP can run a Network Address Translation gateway and > > Why not use net 10 and leave the NATs in? > You can't possibly be serious here. No way NATs scale to multihomed, and they don't speak BGP. They aren't heavy-duty backbone routers. They're designed for small to medium sized corporate LANs, not ISP's. > >Of course, there is one little problem with this.... > >bash$ whois 65 > >Air Force Logistics Command (ASN-LOGNET) LOGNET-AS 65 > >bash$ whois 96 > >Army Finance and Accounting Office (ASN-JTELS) JTELS-BEN1-AS 96 > >How did these guys get such big chunks of address space reserved? > > You're probably gonna regret that one: you're looking at AS numbers, not > networks. Now, let see how many people tell you the same thing :-). > :-) > >I think that the fundamental problem here is that the Internic is > >fundamentally clueless about important issues such as global routing > > Bullshit. The InterNIC is very much aware of global routing issues. > I'm not 100% convinced of this. The InterNIC is somewhat unresponsive to startup providers who are starting multi-homed. It took me almost two weeks to convince them that I needed routable space, and I had to allocate a bunch of networks to various internal uses to do it. Even though I already had bilateral peering agreements at MAE West with about 10 other providers, including several of the big ones. Currently, there are only three major providers that we know of at MAE west that we do not peer with. Two are a matter of time, and one is a policy issue. (on their side) > >and *BUSINESS* issues. > > What business issues are you talking about? > The only BUSINESS issues the InterNIC really seems to be aware of from what I have seen are: Lining the pockets of the gray men Ridiculous attempts to map trademarks onto DNS space > >They are behaving a lot like a government bureaucracy > >or a regulatory agency. > > The registries are simply following the policy as defined by the > Internet community at large. Read RFC 1466, 1814, and the latest > internet draft on allocation policy. If the Internet community were > to define new policies, the registries would implement those as well. > Currently, there is a small bit of contradiction in what small ISPs > and end users want and what the larger ISPs (and what is necessary to > keep the Internet from partitioning) want. But, both sides are more > than happy to scream and whine at the registries for not doing the > "right" thing. > Yes and no. The InterNIC changes policy, templates, and procedures quite often over the last 3 years, often without notice, and enforces the new policies quite rudely, even if they contradict the last policy you received from them. Example: Submit template to update domain Get response back saying "This domain is already allocated" Make phone call explaining that template was an UPDATE, as stated on template Get response saying "Then don't submit a template, just send us a note." Send note Get response "Updates complete" Next weeK, Submit another update note for a different domain Get a response back saying "Updates require submission of a template" ARRRGGGGGHHHHH!! There are REAL problems with the InterNIC, and we need to address them. Is the InterNIC doing an OK job, by and large yes. Are there things wrong, YES. Are the things that are wrong the InterNIC's fault or the policy InterNIC was given? Some of each. > >I don't really see how this can be fixed with the current system of > >having a US government agency writing a contract with a private US > >company to provide a fundamental international infrastructure service! > > The US government agency could get out of the way, but the squeals of > outrage when InterNIC started charging US $50/year for domain name > registrations leads me to believe it will be a while before any sort of > rational allocation policy can be imposed. > I don't see how having the US government fund the agency really has much of an impact here. The NSF has had very little to say about HOW things are done, and merely sets the funding level. The decisions about HOW things are done are made largely by the IANA, IAB, and IETF. > Regards, > -drc > > > Owen ----- End Included Message -----

3 2

Re: Allocation of IP Addresses
by doug＠pbi.net 14 Mar '96

14 Mar '96

> > Can you grow an infinite amount of additional food? Yes, you can. That is how people have been able to eat since the beginning of time. > Can you buy more gold than there is on the planet? We don't know exactly how much gold there is on the planet. That's why we still mine it. > > Can you make more land than we have? Sure, keep on stacking. The point of all this debate from what I see is its dangerous to give up an independent organization which allocates addresses based on need, and go with a cash-on-delivery method. Look whats happening with people buying up domain names just to make a buck reselling them to suckers! These same people would attempt to do the same with addresses. -doug

2 1

RE: Allocation of IP Addresses
by Jim Browning 14 Mar '96

14 Mar '96

It has been suggested that this discussion should be held only on com-priv, as it is global in nature. I am responding to the nanog list only because it is the source of the most recent post.... >From: Andrew Smith[SMTP:awsmith@rip.ops.neosoft.com] > >> >> How many times is it reasonable to ask a customer to renumber? Once is >> >> certainly reasonable. Twice is questionable. More than that and I would >> >> suspect the customer would renumber all right, but as part of shifting to a >> >> different ISP. > >Please tell me how this could happen? I would call this poor planning >on the part of the ISP. Simply request network space from your provider >until you can justify say a /18-16...(if you get a /18, request that >the rest of the /16 is reserved for you and BACK IT UP with documentation) >by noting that you will return ALL of your provider assigned networks >within X amount of time of recieving the delegation from the NIC. This (reserving the /16) sounds like wonderful approach. In our particular case, we promised InterNIC that, if they would reserve a /16 to back up our /18, we would renumber all our customers into that space, returning a directly allocated /19 and two different provider assigned blocks. We are in the process of establishing connections into the NAPs, so having a sufficiently large *contiguous* space is of course very important to us if we are to successfully peer. We invested significant effort in providing what we feel is very adequate documentation. So far, InterNIC has only said that they 'would consider' reserving additional space. Should we proceed with renumbering our customers into a space which *might* have contiguous space reserved??? >This is 1 renumbering. (Not including switching providers if you >so choose...a cost of doing business that you must accept if >it's your choice..although I'll stick my neck out and say this should >happen early when it is least painful...you can usually tell if >you want to leave an provider early on) We are doing our best to avoid 'poor planning', and that is what has created this issue for us. Multiple renumbering will occur only (1) if InterNIC insists on our accepting incremental additions of noncontiguous /18s, or (2) if each of our "small steps" is into a different address space. In the case of (1) we would ultimately have to consolidate these to enable aggregation of our routes. In the case of (2), we would have to constantly renumber. -- Jim Browning

1 0

Re: Allocation of IP Addresses
by Brian Brown 14 Mar '96

14 Mar '96

Perry, You forget one thing. The deep pocket companies CAN buy up all the food or rent ALL the apartments but I can grow more or build more starting today. Show me how I can create more ip numbers today. Brian > > >"David R. Conrad" writes: >> Your proposal is? If you say charging for address space, please >> explain what would stop deep pocket companies from buying up all the >> address space? > >If you start charging for food, what would stop deep pocket companies >from buying up all the food? > >If you start charging rent for apartments, what would stop deep pocket >companies from renting all the apartments? > >Perry > > Brian Brown 703-709-7500 bbrown(a)digex.net

2 1

Re: Allocation of IP Addresses
by Scott Mace 14 Mar '96

14 Mar '96

> Here's an interesting one: > > crash % whois 166.128.0.0 > GTE Telecommunications Products and Services (NETBLK-CDPD-B) > 245 Perimeter Center Parkway > Atlanta, GA 30346 > > Netname: NETBLK-CDPD-B > Netblock: 166.128.0.0 - 166.255.0.0 > > Coordinator: > Brenner, Robert (RB30) ROBERT.W.BRENNER(a)GTE.SPRINT.COM > (404) 391-8450 > > Domain System inverse mapping provided by: > > NIC.CDPD.NET 198.41.0.20 > NETMAN1.NETSOL.COM 192.153.247.11 > > Record last updated on 18-May-95. > > The InterNIC Registration Services Host contains ONLY Internet Information > (Networks, ASN's, Domains, and POC's). > Please use the whois server at nic.ddn.mil for MILNET Information. > This was meant to have the following here: Its interesting why @home didn't get something similar to this... Scott -- smace(a)neosoft.com - KC5NUA - Scott Mace - Network Engineer - Neosoft Inc. Any opinions expressed are mine.

1 0

Re: Allocation of IP Addresses
by Paul Ferguson 14 Mar '96

14 Mar '96

At 09:36 AM 3/14/96 -0800, Owen DeLong wrote: >> >> Why not use net 10 and leave the NATs in? >> >You can't possibly be serious here. No way NATs scale to multihomed, and they >don't speak BGP. They aren't heavy-duty backbone routers. They're designed >for small to medium sized corporate LANs, not ISP's. > Owen, you're assuming that what is currently the case will continue to hold true. - paul

1 0

Re: Allocation of IP Addresses
by Paul Ferguson 14 Mar '96

14 Mar '96

At 11:20 AM 3/14/96 -0600, Scott Mace wrote: > >Renumbering is a fact of life, so live with it and find ways to do it >better and more efficiently. > Thank you for echoing conventional wisdom. :-) I would suggest that anyone with an opinion on this issue [renumbering] obtain and read RFC-1900 ['Renumbering Needs Work'], as well as RFC-1916 ['Enterprise Renumbering: Experience and Information Solicitation'], before firing a salvo over the bow. Renumbering may soon be a fact of life. The PIER [Procedures for Internet/Enterprise Renumbering] IETF Working Group is currently working on documenting renumbering issues to assist organizations in operational aspects of renumbering. Get involved, if this issue concerns you. -paul

1 0

Re: Allocation of IP Addresses
by Paul Ferguson 14 Mar '96

14 Mar '96

I've moved my response and this thread to CIDRD, since it appears to be more appropriate in this venue. - paul

1 0