Well, if the network isn't connected to the Internet, and isn't going to be, then you can use whatever addresses you like. Not recommended to use any old class A (just because requirements always change - you may have to merge with another network, you may need Internet connectivity in the future etc), but it will work. You can use NAT (Network Address Translation) or PAT (Port Address Translation) to translate your 'inside' network addresses (i.e. anything you choose) to legal assigned Internet addresses. This can be done either statically or dynamically. Not all protocols work with NAT; for example if the IP address is contained in the data portion of the application packet, NAT needs to be aware of that and needs to change it there as well (which may mean re-calculating checksums etc - messy). It is generally recommended (actually, I think it is universally recommended) that if you have a choice, use RFC1918 Private addresses if you're not using 'real' (i.e. allocated to you) addresses. These are address blocks set aside and not allocated to anybody. 10.0.0.0/8 is the class A address defined as a 'private' address block. Why might you use some other class A? Well, we use multiple unassigned class As in our internal network. It makes it easier to have an addressing scheme that uses the octet boundaries - you don't need to be able to think in binary to be able to work out what office a particular address refers to. On its own, that isn't a very good reason - but our network addressing scheme was implemented before RFC1918 came out, and it's a large, geographically distributed network with several hundred remote sites (very remote, some of them). The effort and pain that would be involved with re-addressing is, so far, greater than the effort and pain in working around the problems caused by using 'illegal' addresses. Sooner or later we're probably going to have to readdress, but personally I'm hoping somebody comes up with some other solution ;-) JMcL ---------------------- Forwarded by Jenny Mcleod/NSO/CSDA on 16/01/2001 12:37 pm --------------------------- Jennifer Cribbs <[EMAIL PROTECTED]>@groupstudy.com on 16/01/2001 10:21:57 am Please respond to Jennifer Cribbs <[EMAIL PROTECTED]> Sent by: [EMAIL PROTECTED] To: "John Pusledzki" <[EMAIL PROTECTED]> cc: [EMAIL PROTECTED] Subject: RE: subnetting and tcp/ip and Private addressing Help! There is such a thing as private addressing where you can have any address you want, but you usually have something set up at the router that dynamically assigns you an address space when you leave your network, such as out in the internet world. That way you remain legal... And the addressing within your network is your own business then. I do not know how this actually works, but I do know this is an option. I think I read something about NAT providing this service. The group would know about this. I do not. Chuck is a good one to ask on this issue... The question is how come someone can have a class a address that internic did not assign. Anybody?? Jen ------------------------------------------------------------ Have a Good Day!! Jennifer Cribbs [EMAIL PROTECTED] _________________________________ FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED] _________________________________ FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
