1: A LAN is a series of computers connected together with high-speed copper (or, optical, in some very rare cases). Typical LAN technologies include Ethernet, Token Ring, etc. Generally, a single LAN exists at ONE site - One office, one floor, one building, one series of buildings with high-speed interconnects, etc.
A WAN is when you use a circuit or some kind of 3rd party's resources to send your data over a distance between LAN's. The classic WAN is two offices in different parts of town, connected by a T1. Typically, communication across a WAN is much slower speed than a LAN.
When you say "pushes traffic across a demarc into another network", you're getting closer. A Demarc is just a place in a building where circuits terminate, really just a physical place, not anything more complex. So, in a WAN you ARE pushing data through the demarc, but that's part of using a circuit.
The "into another network" part is actually true. I've built LANS with 5,000 switch ports across four buildings. We used about two dozen different network segments (also called separate IP subnets) to manage traffic in the LAN. So, the exit point from a LAN segment isn't always a WAN - It could be another LAN segment.
2: A layer 2 device is charged with transmitting data between two devices on the network. It doesn't care about anything but sending this frame of data from Device A to Device B. A Cisco router has some layer2 components, but is, in general, a layer 3 device as it does routing between networks and works at the IP level. The classic Layer 2 device that everyone uses for a reference is a Bridge or a Switch. They don't care about networks, they just look at the MAC address to see how a frame should be delivered.
Here's a concept for you.. If you have a device that works at layer 3 (such as a Router), it has to also have functions for Layer 2 and depends on Layer 1. For example, a Cisco router is a layer 3 device because it can examine data at the network layer and determine if it should be routed. Once it determines it SHOULD be routed, it functions like any other layer 2 device on the network and transmits across the network.
In general, you really don't separate out the layers when dealing with devices, other than network gear. PC's, servers, etc. all deal at all seven layers, each layer depending on the one underneath it to get data from an application across the network. A router just deals with layer 3 and below, and a switch just deals with layers 2 and below.
A CSU/DSU is a layer 1 device, yes. It is responsable for transferring the physical electrical signal carrying your network traffic across the wire.
3: A gateway is any device that gets traffic OUT of your network segment or IP subnet. A gateway is pretty much always a router. If you have a single building with two subnets and a single router with two Ethernet ports between them, that router is the gateway for both networks. In this case, a 1605 could do the job, as it's got two Ethernet ports. In the case of a smaller network with just one subnet, a 1601 connecting to your ISP could be the gateway.
4: There's no real minimum number of routers you need - If your network doesn't change much and you only have a few routers, static routing is fine. I'd use the number tenish to put in a routing protocol. The time when you would want a routing protocol with less is when you have a meshed network. Example would be three offices with T1's between them. A->B, B->C, C->A. If your A->B link went down, you could still get from A to B through office C, just not the most efficient path. You'd need a routing protocol to realize that a circuit was down and to authomatically re-route your traffic a different way.
- G
A CSU/DSU is inde
Facebook
Twitter