Welcome back to Acing the A+, FSET’s guide to CompTIA A+ Certification! Even if you’re not planning on becoming an IT professional, this blog will provide you with a critical leg up when it comes to the world of computers and Information Technology. Today, we’ll be taking a look at switches, routers and other kinds of network devices.
Ask anyone who works in a data centre, and they’ll tell you there are many different types systems that are used to communicate over a network. Ethernet switches, for example, use cables to connect to local area networks, while routers can split these connections and turn them into wireless Wi-Fi signals. Whether you’re gunning for a job in the field, or just brushing up to better connect your gaming consoles, knowing the difference between these systems is a smart idea.
A router is a piece of hardware that directs and forwards signals from different IP addresses, kind of like an internet traffic cop. Routers will examine the packets of data being sent on the network to determine what’s going where, and how to get each signal to its final destination.
Routers are often referred to as layer 3 devices because what they do takes place at layer 3 – the network layer – of the Open Systems Interconnection (OSI) model. Notably, many kinds of routers can connect and centralize several different kinds of interfaces, ranging from local area networks (LAN), to wide area networks (WAN), to fibre connections all in one place.
If you’re using a copper cable to plug into a computer, you’re plugging directly into a switch. After a router has forwarded traffic, switches keep the signal in motion based on the packet’s Media Access Control (MAC) address. Because of their ‘built in’ nature, switches do their job very quickly – in fact, many switches have an Application Specific Integrated Circuit (ASIC), which is designed to speed up their rate of transfer for up to hundreds of interfaces.
On the topic of switch design, many also feature Power Over Ethernet (POE) technology, which will add additional power to the connection for even higher throughfare. It’s also important to note that there’s a lot – or a little – that can go into a switch. In other words, you might encounter some that have been configured to connect a wide variety of devices on a network, while others might be unmanaged and capable of connecting only a few designated protocols or devices. While you might think using the Simple Network Management Protocol (SNMP) to solve this issue, unmanaged switches commonly have no SNMP capabilities. Regardless, like routers, switches are a kind of layer 3 device on the OSI spectrum.
If you’re tasked with getting switches for an enterprise, unmanaged switches will be less expensive, but managed switches will come with more capabilities – the kinds that help desk staff and system administrators will thank you for. For example, managed switches will enable you to configure different IP subnets and create virtual local area networks (VLANS), as well as prioritize network traffic and throttle connection speeds as necessary.
Spanning Tree Protocol (STP) is often used to organize and prevent loops between switches when there are many connection to one network. With managed switches, it’s also possible to setup port mirroring to help streamline and monitor data packets. When combined with SNMP, this is a use technique when it comes to troubleshooting a network.
Speaking of larger organizations, there will usually be a lot of wires running through the office building. Every worker’s computer will be wired to a central wiring closet (or several if it’s a multi-floor complex), unless they are able to connect to a wireless access point, which will likely be mounted to the ceiling. Within the wiring closet, every ethernet cable will be hooked up to a switch with an RJ45 connector. Buildings with infrastructure that is setup properly will be able to accommodate team moves and workstation movement with ease, simply by moving the necessary cables around at the wiring closet.
Regardless of the size of a network, in this day and age it’s likely that it will have a firewall. Firewalls allow and disallow traffic over a network based on the IP addresses and port numbers attached to the travelling packets of data. Because firewalls track TCP and UDP ports – which are themselves transport layers on the OSI models – it is fair to say that they are a layer 4 device (or even layer 7 if they understand application layer traffic!). As far as software goes, firewalls are quite nifty and present with a lot of functionality; they can serve as endpoints for encrypted communication, routers, and even as proxies, kind of like a browsers-within-a-browser that can function like a barrier between users and malicious code.
Nowadays, companies cut back on cabling when and wherever possible. Using POE, many devices, ranging from access points to security cameras, are powered with ethernet cables plugged into their switches – which is known as an endspan. If a device’s switch doesn’t support POE, then you’ll need to modify the ethernet so that power is injected in the middle of the connection. This is a midspan. Notably, most contemporary switches are marked with what they can and can’t connect to; if you see a switch with a blue top, it will support POE.
It’s important to note that different devices require different kinds of power, and that there are respective POE types and standards. The traditional style of POE is IEEE 802.3af, which in the years since its inception has since been incorporated into the standard 802.3 ethernet cable – as has the upgraded IEEE 82.380at variant.
Another modern variant of POE is IEEE 802.3bt, which is referred to as a Type 3. Type 4 POE, meanwhile, can provide up to 71.3 watts of power with 960 milliamps, meaning its wired for 10 gigabit per second connections. POE will continue to be upgraded well into the future, so long as ethernet cables do not become obsolete.
Before switches, the industry standard for connections were hubs, also known as multi-port repeater. While switches are good at differentiating and compartmenting data, hubs simply can’t, so you can probably understand why they’re obsolete – in fact, the more connections a hub had, the slower it would run, and with such slow speeds to begin with, it was never a good thing.
Do you have a cable modem in your home, hooked up to your television? These kinds of cables can send video signals, phone signals and internet signals, all thanks to the Data Over Cable Service Integration Specification (DOCSIS) standard they’re based on. On the other hand, your telecommunications company might have provided you with a DSL modem – short for Digital Subscriber Line – which does internet signals only. Sometimes these modems are ‘asymmetric,’ meaning that their download speed is significantly faster than their upload speed, by design. There are limits on how far DSL modems will work away from the telco’s central office, and they will generally work faster the closer they are.
Last but not least, fibre connections are becoming increasingly prominent on the market. Fibre connections use Optical Network Terminals, which are usually setup outside of the building in question. In the industry, these outdoor terminals are referred to as a demarc because they demarcate and delineate data from your Internet service provider (ISP), meaning that the cabling feeding into it is their responsibility, and the cabling coming out of it is yours.
Finally, all of the wired connections discussed thus far involve a Network Interface Card (NIC). NICs are part of any given device’s circuitry; sometimes they are standalone chips, like a graphics card, and sometimes they be built in as a part of the device’s motherboard. Laptops, computers, routers, and everything else that can receive a wired connection will have a NIC.