Table of Contents
We have covered the theory of LAN's and WAN's earlier in this course. Now we need to cover the connection to an ISP, which constitutes the latter.
Running TCP/IP over a LAN and a WAN require different encapsulation protocols. What this means is that when communicating over a WAN, delay needs to be factored into the communications, handshaking, signaling, etc. LAN's of course have all this built-in in the TCP/IP protocol already. For this reason, TCP/IP can be transported over a SLIP link or a PPP link. SLIP stands for serial line IP, while PPP stands for Point-to-Point protocol. SLIP was the predecessor of PPP and did a good job, but for some reason PPP gained acceptance and is now the de-facto protocol for transmission over serial links.
What happens when we send packets from a LAN to the WAN? When the packet reaches the router (in this case the router is the same machine as our Linux machine), it is "translated" into PPP packets and sent over the wire. Since communication is slower over a WAN, time-to-live settings need adjustment, as well as sequencing of packets from source to destination and vice versa.
It is difficult in a course such as this to describe all possible problems that one needs to deal with when configuring the WAN, but I am providing a checklist of those that may occur and possible solutions to them. The order of this checklist is important.
You will need to know a number of things from your ISP, so here is a checklist before you begin:
What is the ISP's telephone number for their dial-in subscribers? (dur)
What type of authentication do they use (PAP or CHAP).
Your username and password for authentication.
What is their DNS server's IP addresses?(there should be two of these)
What is their SMTP server's IP address (or name)?
Who can you call with technical queries?
Do they provide you with an IP address dynamically or do you need a static one from them?
Armed with these answers, you can begin the following checklist.