Version 4.1 - 1.6 Tunneling Protocols
IP Tunneling is a topic that I have always found more difficult to understand or explain as a concept than it is to configure. Usually the concepts are easy and the configuration complex.
Basically, you create a tunnel interface and configure it for the specific type of tunnel you want. Logically, you have to tell it the source and (optionally) destination. Then you have to tell it what traffic you want to put into the tunnel. [This will require an access-list.] As long as you already have a route on the source to the destination (and vice versa), you're good.
[The biggest issue with tunnels is recursive routing which happens when you try to route to the destination using the tunnel itself. ]
But this is like trying to explain what "shoes" are to a kid. It's a little overly simplistic when you speak about it in general terms. You show them a pair of shoes - perhaps an oxford. Shoes cover your feet. Are sneakers still shoes? What about sandals? And pumps...
It's a silly little example, but the point is that there are a lot of types of tunnels. I don't think all of them will be on the exam. The blueprint subdivides this into three topics: GRE tunnels, NHRP and IPv6 tunnel types. There are so many others - some of which are actually covered later under other topics (like VPNs).
Some study guides try to break tunnels into three components:
Passenger protocol
Carrier protocol
Transport protocol
Think of a person on a train. The person is a passenger - the item that is going from source to destination. This is whatever traffic you're sending through the tunnel (like IPv6 traffic). The carrier is the train. This is whatever you're wrapping around the passenger protocol. GRE, IPsec - these are carrier protocols. And the transport protocol is the track itself. This would be IP.
I think a lot of this is more oriented towards the R/S track. But you see how uneasy it is to speak about tunnels in general. So let's move on and talk about specifics.