if you have some form of broaband access like ADSL the provider typically uses the PPP protocol framework.
in this case PPP over ATM or PPP over Ethernet or PPP over ethernet over ATM.
the PPP session is point-to-point but multiprotocol capable.
Because it has a protocol type field in its header.
Once a basic PPP session is established possible L3 PPP protocols are used.
IPCP is the PPP auxiliary protocol that manages the assignment of the IP address.
On a Cisco router you can see the whole process using
debug ppp negotiation
debug ppp authentication
after successful L2 negotiation and authentication IPCP PPP frames messages are exchanged and the provider takes an address from one address pool and proposes it to the home router.
The home accepts it and sends a positive acknowledge (CONFACk : accepts the configuration proposed)
the home router needs to have
ip address negotiated
in sh interface dialer means the IP negotiation has been successful and an IP pipe is ready to be used.
if you use a cable modem, DHCP protocol with specific extensions for cable modem scenario is used to assign an ip address.
for some business connection, DSL lines have a static ip address assigned and in this case the provider tells what ip address has to be configured manually on the user side.
Hope to help
Not sure if you are referring to the Internet address assignment
Each ISP would have a set of IP pools (mostly a supernet) available with them registered with the IRR. ISPs subnet it further and allocates the ip address based on the customers requirement and route it towards the customer (this could either be via statics or might be learnt via BGP.
These addresses would then be summarised before it is passed on to the peers or NAP
Narayan thanks for the reply.U got me correct i was refering to this.
I am confused about the process of this assignment and mainly about classless ip addressing.
I have a good understanding for the ip addressing but i think i m missing something to get the concept of this.
Can u explain me the proceess with a example.
say for eg an ISP has been alloted the following subnet 100.100.0.0/16 from the IRR. THe ISP will subnet this adddress and allocate as required.
Say Customer A may require 15 odd ip address while customer B requires 250 address.
ISP will assign 100.100.10.0/27 to CustA
and 100.100.11.0/24 to CustB. Due to longest prefix, the routing within the ISP cloud would ensure that the traffic is routed to the correct destination.
But it does not make any sense to advertise these individual preixes to the upstream ISP (infact no ISP allows you advertise anything less than a /24). So the ISP would summarize the netowrk and advertise it as a /16. Traffic destined to these address will be routed to the ISP which will then internally route to the correct destination based on the longest prefix. This minimises the number of routing prefixes advertised
HI Narayan thanks for u r reply.
As u said that the IRR alloted the subnet 100.100.0.0/16 as i understand from this the address given by IRR is already subnet with class B mask but actually its a class A address m i right in thinking this.
and as u said in the example of 100.100.10.0/27 for CustA what will happen to the rest of the IP it that they will allocated some other cust with the similar needs..?
So u mean to say that this subnetting is local to the ISP and why does the ISP's dnot allow anything less than /24.
Yes 100.100.0.0/16 is a subnetted class A address.
Rest of the IP will be alloted to some other customer.
ISP can allot /27 address to 8 customers (makes a /24 address), instead of advertising /27 address for each customer
ISP advertise /24 to reduce the routing table.
Thanks for your reply.
I understand that /27 will give u 8 subnets but will it make a /24 address can u please explain.
And if the provider advertises /24 out to others how will the outside networks reach them as they are subnetted address.
100.100.10.0/24 can have following /27 address,
when you summarize these 8 address you will get 100.100.10.0/24 network.
Let say 100.100.10.34 which is 100.100.10.32/27 customer address will reach this ISP router who is advertizing 100.100.10.0/24, this router know that this address belong to customer 100.100.10.32/27 .
Thanks for your reply.
I got how summarization is working here.
If u take the example of subnet allotted by IRR ie.100.100.10.0 suppose 100.100.0.0 is assigned by IRR to the ISP,then ISP is again going to subnet it and assign it to customers thats fine.
But if u analyze the subnet allotted by IRR is a class A network now my question is a the TOP level i mean at the NAP level how will this subnet advertised.
customer(100.100.32.0)ISP(100.100.00)how about NAP how will it advertise to other NAPS is it as a class a network as 184.108.40.206 or something else.
It will be advertised as the supernets only by the NAP to other NAPs..
The NAP is just aware that the class A address is originated by the ISP AS nad has no knowledge of the subnetting inside the AS
Narayan thanks for the reply.
So u mean to say that it will be advertised as a class A address..please confrim.
For my understanding is the topology flows like below.
subsciber(100.100.10.0)--->ISP(100.100.0.0)-----TOP Level Provider(220.127.116.11)-------NAP.m i thinking this correct or i m wrong.
Nope it will sent as 100.100.0.0/16 unless all other networks from that range are subtended from the same NAP.. it can then summarize them further and pass on
Thanks for the reply.
If they are from the same NAP they will be advertised with 18.104.22.168/8 or /16..?.
If the some subnets from the same address space are allotted to some other providers which connected to a different NAP then what will be the routing advertisement and routing table would look like.