Subtitles of the Movie

When we talk about switches, most of the time we are going to be talking about layer two switches, which are in effect fancy bridges. Switches will eventually replace bridges, because they have more ports, they work faster, they give the network administrator more flexibility, they support virtual local area networks, VLANs. But they do work at layer two; layer two switches work at the layer two of the OSI model. But you may also hear the term layer three switch. Well, a layer three switch is basically a programmable router, because if it works at layer three, understand that is the network layer, if it is working at the network layer, that means that it's using a layer three logical address. So if it's using a layer three logical address then it is routing; if we can program it, then what we end up with is a programmable router. But layer two switches work very much the same as the illustration that we just did with the bridge, they learn MAC addresses. But switches can do it faster and can move information much faster than bridges can. Switches are hardware that is specifically developed to create collision domains by learning those MAC addresses. And a ten megabits per second switch has a bandwidth, or can support a bandwidth of ten megabits per second on each port; whereas a ten megabits per second bridge divides bandwidth, so switches make much more efficient use of bandwidth than bridges do, and that's another reason that switches are replacing bridges. Later we'll talk about virtual local area networks, but we certainly cannot set up a virtual local area network without using a switch. So switches separate a network into multiple collision domains, that's layer two switches, layer three can also separate a network into multiple broadcast domains, because as we said, they are basically programmable routers. For the rest of this training, when I say switch, I am really talking about a layer two switch, because a layer three switch in essence is a router. But it is important to understand that switches are an essential part of a network topology, it's an essential device to create multiple collision domains. In this way we are segmenting our traffic; so if I have two different devices that work off of the same port on a switch, those devices can send large amounts of data to and from each other, without bothering any other segment of the network. And that is the main advantage of switching. I can segment network traffic so that the traffic doesn't even go any further than one segment; in other words, it goes to the switch and the switch says - well, you know what, if your destination is the same as your source in regards to the port, in other words. If you are coming in my Ethernet one port, and you are looking for something that I know is in my Ethernet one port, well then there is no point in me sending you any further than I already have, and there is no point in me disturbing anybody else in this network. So I can have two devices that are in the same segment, send tremendous amounts of information back and forth to each other without creating any other traffic in any other segment. And that is the main advantage of switching; now when we have multiple switches, we have multiple advantages, but we can also run into multiple problems because of the way that switches flood. So we need to use a protocol developed by Cisco called Spanning Tree Protocol. In our next section, we'll discuss the Spanning Tree Protocol. That's next. alled Spanning Tree Protocol. In our next section, we'll discuss the Spanning Tree Protocol. That's next.