On Mon, Feb 7, 2011 at 6:22 PM, James Hill <[email protected]> wrote: >> ... NAT is expensive, not just in terms of equipment but in design >> and installation costs, and the tech support needed for all the >> things that don't work behind NAT. ... >> ... freeing small blocks of address space here and there >> doesn't really help the big picture, but does grow the size of the >> global routing table significantly >> So it seems like several high costs, with little practical gain. > > It would not have been as high a cost if it was done much earlier.
How so? (You make a statement but do not provide a supporting argument.) >> Apples and oranges. The closer you get to the default free zone, the less >> operators care about individual nodes or small nets; it's all about getting >> packets to the next large operator. > > I have already provided an example of a very large network that is > successfully issuing private IP's to hundreds of thousands of customers. By "private", I presume you mean addresses from the RFC-1918 blocks. What connectivity exists between these endpoints and the outside world? Are they just using these addresses for management or VoIP, thus completely alleviating the need for NAT or anything like it? Are they using WAP and friends, or HTTP proxies, or web accelerators, thus not needing outside connectivity at layer 3? > I don't need to know the specifics when I can see that it is possible. But as far as I can tell, you don't actually know what's realistically possible. If you know more than you've stated, please inform us. So far, all we seem to know is that "Telstra is large" and "Telstra uses RFC-1918 addresses". Comcast uses RFC-1918 addresses, too. They use them for internal management purposes; packets with those addresses should never leave their network. There is zero NAT or outside connectivity involved. > I never suggested that "anyone" can do it but it is reasonable to expect that > a university, for example, with all of its resources can find a way. When I worked for IT at a university, we always had very tight budget rules. You cannot hand wave the problem away by saying "Universities are large and thus have lots of resources". They have lots of resources because they are large; that doesn't mean they have *excess* resources. >>> Can you even articulate who it is that is in need of these addresses but >>> can't >>> get them? >> >> How is this also not blindingly obvious? ANYONE that needs an address. >> >> You seem to be answering the previous poster's question with "No". > > An answer of "anyone" does not equal "No". I'm not sure how that can be any > clearer. Repeatedly stating things like "blindingly obvious" and "I'm not sure how that can be any clearer" does not increase anyone's understanding. To break this down into discrete statements: A1. You claim universities don't need all of the addresses they have been assigned. A2. Others ask you for who should be able to obtain these unneeded addresses. A3. You say "ANYONE who needs" one should be able to obtain an address. A4. The set "ANYONE" includes "Universities". For the case of an organization which wants to voluntarily give up part of their address space: They can do so any time they want. The issuing Registry will happily accept such. Unsurprisingly, organizations which have done so have been few. For the case of an organization which have addresses without a node currently connected: Balancing resource allocation between current need and future expansion is the fundamental theme of this discussion. Your argument seems to be that we should (be|have been) more conservative when handing out address space, because there will likely be need for it in the future. If I'm the organization, why shouldn't I adopt that argument for my currently allocated address space? Relatively recently, the idea of assigning every dorm room multiple IP addresses seemed unlikely. Today it's common-place. Huge increase in address space usage there. >> Define "needs an address". You're saying the univ's don't "need" >> their address space. So who does? What's the criteria? > > Now who is generalising? I did not say uni's don't need there address space. > I said that there appears to be many that don't need all of it ... It appears we have a miscommunication. Allow me to restate: Define "needs an address". You're saying the univ's don't "need" all the address space allocated to them. So who does? What's the criteria? -- Ben ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~ --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to [email protected] with the body: unsubscribe ntsysadmin
