Ok, im getting fairly desperate now.

I am banging my head against the wall trying to learn the ins and outs of the TCP/IP protocol.

ive figured out how to assign a subnet to an IP and how to figure it out what is needed depending on how many hosts per network i want

BUT

I cant figure out all the other supernetting stuff, its just baffling me, i get lost reading it.

can anyone help me talk to me? if you are in the UK i will telephone you and we can talk over it, im so desperate its unbelieveable.

if you want me to give you a hosting account for free i will, domain, e-mail everything, 10 gig transfer, all for free, a few hundred meg of space for free, i need to figure it out.

PLEASE HELP ILL BE ETERNALLY GRATEFUL

Hey, I am doing my MCSE course subnetting and everything.. maybe I can help you out but I am not in the uk...

You can get a subnet calculator at http://solarwinds.com..

What don't you get...:eek:

BGP 4 supports classless interdomain routing (CIDR), which lets you reduce the size of your routing tables by aggregating routes, resulting in supernets. CIDR eliminates the concept of network classes within BGP and supports the advertising of IP prefixes. Basically a supernet is just a lot of network advertised as one.
For example I can advertise 208.155.64.0/24 (A C block) and 208.155.64.0/24 (Also a C) as two routes. Or I can advertise from
208.155.64.0 all the way to 208.155.79.255 with one supernet
208.155.64.0/20

Also Cisco has a decent article on subnets at
http://www.cisco.com/warp/public/701/3.html

and you might want to look for a document floating around the net called daryl's or darly's (I forget which) TCP primer. It is very condensed but also very good and it might help you get a hold on it. If there are any specifics you'd like or if you'd let us know what your trying to do I might be able to help.

damn , i never expected responses that quickly.
thanx people :)

Kanwarpal: thanx for the subnet calc, but im really looking for the theory behind it all...

smoats - thats a very good explanation of it makes a little more sense of it now.

its not for a specific configuration ots more a knowledge thing, as part of my job ive got to be able to explain it in simple terms just like smoats just has, i could do with sitting down with you for a while smoats, would you like to have a chat sometime after work over netmeeting ?

if not i understand, a guy like you must be busy, lol

another problem i am having is understanding Network ID's

for example:

IP Address : 129.56.189.41

Subnet : 255.255.240.0

Network ID: 129.56.176.0

WHY is that the network ID is it the start of something? am i misunderstanding it ? WHY is it that number and what importance/meaning is that number?

First, subnets/supernets are not really a TCP/IP issue. TCP says nothing at all (it leaves datagram delivery up to IP), while IP merely assumes that there is some routing mechanism.

But on the topic of routing: smoats is correct about CIDR/BGP4/route aggregation. People are "encouraged" to aggregate routes, meaning that where they would otherwise advertise two contiguous /24 netblocks they would advertise a single /23; the point being, of course, to keep routing tables at a reasonable size. (Routing table inflation is generally considered to be the largest outstanding problem with IPv4).

To answer xtstrike about network IDs: on many old systems routing was performed following the "longest matching prefix" rule: Route a packet towards the network which has the longest matching prefix. In other words, if you have entries for 1.2.0.0/16 and 1.2.3.0/24, a packet for 1.2.3.4 will be routed towards the second network while a packet for 1.2.4.5 would be routed to the first network (since it doesn't match the second entry). This lead to networks being "identified" by their prefixes, ie network = IP AND [subnet mask], where "AND" is binary of course.

Yep,
Your correct routing is an IP level operation, I was just refering article that was named somebody's TCP primer.

Sam

Originally posted by smoats
Your correct routing is an IP level operation, I was just refering article that was named somebody's TCP primer.


No. Routing is not part of the IP protocol. IP assumes that there is some external mechanism to handle routing.

IP can be run over Avian Carriers using random routing (ie keep on forwarding packets randomly until they reach the correct destination) in complete conformance with the appropriate standards.