It takes many different devices to keep a network running. In this video, you’ll learn about switches, routers, wireless access points, and many other devices that we use on our enterprise networks.
<< Previous Video: Common Network Ports Next: Installing a SOHO Network>>
As an IT professional, you’ll be working with many different networking technologies. So in this video, we’ll look at some of the most common network devices. If you are connecting a device to a network, whether it’s a wired network or a wireless network, it needs some type of hardware to be able to make that connection. We call this piece of hardware a Network Interface Card, or a NIC.
So you will find a NIC inside of your printers and your servers and your laptops and your workstations and anything that needs connectivity to a network. The network interface card that you’ll be using will be specific to the type of network you’re connecting to. So if you have an ethernet network, then you will need an ethernet NIC. If you have a wireless network, you will need a wireless NIC. And if you are connecting to multiple types of network, you will need multiple types of NICs inside of your device.
The network interface card that’s here is an external adapter you would plug into the motherboard, but many motherboards have a network interface card built into the motherboard itself. There are also many other types of network interface cards. This one happens to be a copper ethernet interface, but you can get network interface cards that have fiber connections, Wide Area Network connections, multi-port connections, and other options.
If you’ve ever had to extend a network connection over a very long distance, you know there is a maximum link that is supported for that particular topology. One way that you can extend this link to be even larger is to use something like a repeater. A repeater receives a signal, regenerates it, and then resends that signal out another interface. It doesn’t have to make any forwarding decisions. It doesn’t have to decide which connection this is going to. This is a simple goes in one connection and goes out of another connection.
It’s very common to use these repeaters to extend the length of a network. So we might be extending a fiber network, or we might be extending a copper network. Or it might be the situation we have with this repeater where we’re converting from one physical type to another. For example, we’re coming in at 100 megabit ethernet over fiber, and we’re outputting this repeater at 100 megabit ethernet over copper.
In the early days of networking, if you had to connect a lot of different devices together, you might use something like a hub. This is an ethernet hub. This is a very small ethernet hub with only four interfaces on it. But these hubs could be tens or even hundreds of interfaces in size.
The way that these hubs operate is that information that is sent to one interface on this hub is automatically repeated to every other interface on this hub. This is very similar to the functionality we saw with repeaters, but the repeater was only repeating it out a single interface. With a hub you’re, repeating it out of multiple interfaces simultaneously. You will sometimes hear a hub referred to as a multi-port repeater.
The communications process on a hub all occurs at half-duplex because of this repeating functionality. This means that two devices can’t communicate at the same time on a hub. You can have one device sending traffic. Once that device is done, another device can then begin sending information.
If you don’t have a lot of devices on the network communicating to each other, then this half-duplex functionality is just fine. But as more devices begin communicating, the efficiency of the network begins to decrease. Ethernet hubs only operate at 10 megabits per second or 100 megabits per second. You won’t find any gigabit speed hubs. In fact, it’s hard to find 10 or 100 megabit hubs today, because the technology doesn’t scale as you put more traffic on the network.
In these early networks where we used hubs to connect all of our devices, we would connect the hub networks together by using a bridge. These bridges make decisions on what traffic should be forwarded through the bridge based on the destination MAC address that’s inside of that ethernet frame. That certainly sounds very familiar, because that’s the same type of forwarding decision made by today’s modern switches. But back in the day, these bridges only had two or maybe four interfaces available to be able to make those forwarding decisions based on MAC address.
These bridges allowed us to make networks that were a bit smaller so that each one of these hub networks was able to operate efficiently. And sometimes we would use these bridges to switch between different topologies. So we could move from an ethernet network to a WAN network by sending that traffic through a bridge. Instead of making traffic decisions like a hub which took traffic from one interface and repeated it to all of the other interfaces on that hub, the bridge was a little more intelligent with how it would decide where traffic was going. It would look at the destination MAC address, find out what interface on the bridge that destination MAC address existed, and then would send that traffic to only that interface where it was destined.
A good example of a modern version of a bridge would be a Wireless Access Point where you have a wireless network on one side, and on the other side, it’s connecting to your wired ethernet network. That wireless access point is performing a bridging function. So it’s looking at the destination MAC address of the traffic it receives, and it’s deciding whether it should send it on to the wireless network or whether that traffic should go onto the wired network. These days, we’ve extended this idea of bridging into very large scale systems that have hundreds of ports on them or are making these forwarding decisions in the hardware of these devices. We call these newer style bridges switches, and we’re able to support huge infrastructures with hundreds of devices on a single switch by using this switching technology.
The switches are performing exactly the same function that these bridges did. It’s looking at the destination MAC address, and it’s sending that information to the appropriate interface on that switch. It’s able to do this very, very fast across hundreds of different interfaces by performing this switching look up in hardware. This switching hardware is an Application-Specific Integrated Circuit, or an ASIC. And it’s this hardware switching that allows us to scale this up to hundreds of interfaces on a single switch.
If you look at the core of an enterprise network, you’ll probably see a switch like this with hundreds of interfaces on it, or it may be a smaller switch that’s in a networking closet on another floor or used in a small office or home office. As we mentioned earlier, these switches make their forwarding decisions based on the destination MAC address of the traffic going through the switch. As we’ll find out later in this video, if a device is making its forwarding decision based on the destination IP address of the traffic, then that is a router. There are some switches that allow you to have both switching functionality and routing functionality within the same device. We refer to these as multi-layer switches or layer 3 switches.
If you’re installing a network switch and you need very basic functionality, then you’ll probably want to use an unmanaged switch. There’s not a lot of configuration involved with setting up an unmanaged switch. You simply turn it on, plug in all of the devices, and they can all communicate to each other. There’s usually not even a configuration tool or utility you would use to configure the switch. You simply connect all of the devices, and they would all communicate across the same virtual LAN.
This also, obviously, would not integrate with other external protocols. If you needed this switch to be able to communicate back and forth to a management station via SNMP, then you probably wouldn’t use an unmanaged switch. The trade off, of course, is that the cost of the switch is lower if you don’t have to support all of these other features. So if all that you need is basic connectivity at a low price point, then you may want to consider an unmanaged switch.
Many organizations, though, need additional functionality in their switches. And in those cases, they would purchase and install a managed switch. Managed switches allow you to configure different VLANs on different interfaces, for example. You might also be able to connect switches together in a trunk. You might also hear those referred to as 802.1Q.
There might be traffic prioritization on the switch so you can decide what types of traffic have a higher priority than other types of traffic. There’s also some redundancy support you may be able to configure in a managed switch by using Spanning Tree Protocol. And our network management station can communicate to these devices using a specialized protocol called Simple Network Management Protocol, or SNMP. And for people that need to do troubleshooting on the switch, you can set up port mirrors, so traffic can be mirrored from one port to another. This allows you to connect a network analyzer to one of the ports on the switch and copy traffic from any other port on that switch to watch the traffic flows across the network.
A device that makes forwarding decisions based on a destination IP address is a router. These are usually standalone devices, but sometimes that routing functionality can also be integrated into switches. We usually refer to those as multi-layer switches or layer 3 switches.
It’s also very common to use routers to connect different types of topologies. So we may connect a serial WAN link, an ethernet copper connection, and an ethernet fiber connection all on the same router. Many organizations provide access to wireless networks by using a Wireless Access Point, or a WAP. Although these wireless access points look very similar to the wireless router that you might use at home, the operation of these devices is quite different.
The wireless router, you have at home is not only a wireless device. It’s also a router switch. It has other functionality as well. In comparison, a wireless access point is simply extending a wired network onto a wireless network and allowing connectivity between those topologies. A wireless access point is making its decision based on the destination MAC address. Therefore, it’s acting also as a bridge.
If you walk around a large facility such as a hospital or a university, you’ll notice there are a large number of wireless access points as you move from building to building. And of course, someone has to manage all of these wireless access points on the network. To be able to centralize this management, these organizations use a wireless LAN controller. This allows you to have a central management console to be able to support hundreds or even thousands of wireless access points wherever they happen to be on your network.
If you need to deploy a new access point, change the configuration, update software, or anything else associated with the management of that device, you would use one of these wireless LAN controllers. This is usually a proprietary system. So if you have a Cisco access point, you’re probably using a Cisco wireless LAN controller.
Although we’re showing a physical device here to represent a wireless LAN controller, there is some wireless LAN control software that runs in the cloud. So you can simply connect to the cloud-based controller from anywhere you happen to be able to manage all of those access points on your network. Many organizations use firewalls to be able to manage the control of traffic flows through their network, especially traffic flows that are going to or coming from the internet.
A traditional firewall allows you to filter information based on the UDP port number or the TCP port number. You may sometimes see this referred to as OSI layer 4 filtering. But modern firewalls are able to examine everything in that traffic flow, including the application that’s in use. So a security administrator can tell the firewall to allow database transactions but prevent file transfers through the network.
Many firewalls also rely you to create encrypted tunnels to and from that firewall. So if you’re off site, you still need connectivity to the corporate network. You can connect over a secure channel to the corporate firewall and then be able to communicate to your internal resources. You might also even find older firewalls that act as a proxy, which means that they sit in the middle of the communication.
If you wanted to surf a website, you would send that request to the firewall. The proxy firewall would then make the request for you, receive the response, check through the response, and make sure it’s appropriate for you to view and then send that traffic to you. It’s also common to see many firewalls used as a router. Sometimes you’ll see this referred to as an OSI layer 3 device. These routers are able to also sit on the edge of the network and be able to do any type of routing or network address translation based on the routing engine inside of the firewall.
A common network device on both home and corporate networks are cable modems. These allow you to connect to a broadband network, usually provided by a cable television company, that is sending data across the network using a standard called DOCSIS. That’s Data Over Cable Service Interface Specification. These DOCSIS networks support many different types of throughput. You can have slower networks at four megabits all the way through his higher speed networks at 250 megabits per second. And these days, it’s not uncommon to see gigabit networks running over these cable modem networks.
Another important aspect of these DOCSIS networks is the support for multiple services. We’re already supporting video through this network, and now we’re including the data for our internet connection and telephone communication with voice as well. For both home and business networking, DSL is a viable competitor to the cable modem networks. Instead of using the same cable used for cable television, a DSL network is going to use the same wire that we traditionally use for our telephones. DSL stands for Digital Subscriber Line, and you’ll sometimes hear it referred to as Asymmetric Digital Subscriber Line.
It’s asymmetric because the download speed that you receive on DSL is faster than the upload speed, making it an asymmetric communication. One challenge we find with DSL is there is a significant distance limitation between the telephone company’s central office and the telephone jack that’s inside of your home. The maximum distance you would be able to get on DSL is somewhere around 10,000 feet.
If you live close to the central office, your throughput will be faster than someone who lives farther away from the central office. DSL speeds generally range around 52 megabits per second downstream and 16 megabits per second sending traffic upstream. And if you live closer to the central office, your DSL connection may even be able to support faster speeds than that.
This is a picture of a traditional company’s network configuration on the floor of the building. You’ve got a lot of people out on the floor that are working at their desk, and all of these devices have wires, ethernet cables, that go into the ceiling or under the floor into one of the closets that’s somewhere nearby. And in that closet is a patch panel. This patch panel terminates those wires onto what we call a 110 block and provides, inside of that closet, a set of RJ45 connectors that go all the way back to each person’s individual desk.
Also in that closet, we have our networking equipment. So we’re able to create a simple patch between our patch panel and our networking equipment by simply extending some ethernet patch cables inside of our closet. This patch panel also allows us to make changes very easily. We can simply extend a connection. And if we realized we wanted to connect that user to a different switch, we can simply move the wire down to a different switch.
If they’re on a different VLAN or a different network, then we can connect those users to a different switch or a different set of interfaces on the same switch. This means you only have to punch down all of the wires from everyone’s desk one time. If someone then moves from one desk to another or we need to plug that person into a different type of network connection, we simply disconnect the cable inside of our closet and plug it into the new connection.
We don’t need to run a new cable from someone’s desk. We don’t need to use any special tools inside of our closet. We simply connect and disconnect using the RJ45 interfaces on the patch panel and on our networking equipment.
Traditionally, if you were installing a wireless access point, there would usually be two connections for that wireless access point. One connection was to provide the network connectivity for the access point. The other connection was to provide the power, and then we would need to plug into a power outlet to be able to power that access point. Today, we’re providing that power over the ethernet cable itself. We call this Power over Ethernet, or PoE.
This means that we can run a single cable now to our wireless access points, to our phones, to our security cameras, and we don’t need any additional connections to be able to power those devices. The power that we have on a PoE connection is often coming directly from the switch, and you have a single run all the way to that device. In those scenarios, we call that an endspan. In some cases, you may need to install a device that requires a PoE connection, but your switch does not provide any power.
In those cases, you can put a device in the middle, like this PoE injector, which adds power to the ethernet connection so that you can then power that device. Most switches that support power over ethernet will say it on the switch itself. For example, this is a 10-point gigabit power over ethernet managed switch.
Power over ethernet allows us to power device using our ethernet cables. Ethernet over Power, or EoP, is the reverse of that where we are extending our ethernet network using the power cables that we already have in our home. You might also hear this referred to as PLC or Power Line Communication. And it’s an IEEE standard numbered 1901. This EoP standard operates at 500 megabits per second, and it’s designed to connect devices that normally wouldn’t be connected to our ethernet. Networks for example, if we had an electric car that we recharged overnight, when we plugged it into the power source, it would also be connected to our ethernet network.