CompTIA Network+ N10-008 – OSI Model
Open Systems Interconnection (OSI) Model Overview The OSI model So what is the OSI model? Well, it stands for the Open Systems Interconnection Model, and it was developed back in 1977 by the International Organization for Standardization, or ISO. Now, you’re usually going to hear it referred to as the OSI model. People don’t usually say the Open Systems Interconnection Model, but you will also heart referred to as the OSI stack. And this is a really fundamental thing in networking, and it is extremely important—so important that you’re going…
The OSI model So what is the OSI model? Well, it stands for the Open Systems Interconnection Model, and it was developed back in 1977 by the International Organization for Standardization, or ISO. Now, you’re usually going to hear it referred to as the OSI model. People don’t usually say the Open Systems Interconnection Model, but you will also heart referred to as the OSI stack. And this is a really fundamental thing in networking, and it is extremely important—so important that you’re going to get a lot of questions on the network plus exam about it.
Now, this model is made up of seven layers, and I’m going to talk about each of those seven layers in the next seven videos because this model is just that important. It’s useful when we’re trying to troubleshoot networks because if I have a problem and I can think through it through those seven layers, I can identify where the problem is and then troubleshoot it appropriately. And this model is going to serve as what’s called a reference model. Now, you may be wondering, “What’s the purpose of a reference model?” A reference model, on the other hand, will categories network functions into specific layers, which is what the OSI model does.
In fact, when you start looking at the OSI model and start learning it in depth, you’re going to find out that the OSI model doesn’t easily and accurately refer to the way our networks operate today. But with the OSI model, it does refer to how networks work for all networks, no matter what time period it is. And we’ll talk about that more after we get into the seven layers. Now, the idea of this reference model is that it allows us to compare technologies across different manufacturers.
So if I’m looking at a particular wireless card, I can compare it and how it operates through the seven layers of the OSI model and compare it equivalently to cards from different manufacturers. And by understanding the functions at each and every layer, it really is going to help us better understand the flow of data in the network, how it communicates, and how we can troubleshoot it. So what are those seven layers of the OSI model? They’re here for you on the screen. We start at the bottom and work our way up.
At the bottom, we have the physical layer, then we have the data link, network transport, session, presentation, and application layers. Now, you may be asking yourself, “Why is there a piece of pizza on the screen?” Well, that’s a piece of my favorite pizza, which is sausage pizza. Now, why is sausage pizza important in a networking course? Well, it’s how I remember the OSIModel layers because, on the exam, you’re going to get questions about these layers. They’re going to ask you, “What is the 7th layer of the OSI model?” And you have to say “application.” Or they might say, “What layer is the session layer at?” and you’ll have to say five.
And then you’ll get other, more complex things, like asking about a device and figuring out what device goes in which layer, and questions like that. So we’re going to do a lot of digging into the PSI model. But back to the sausage pizza. Where does that come into play? Well, I like to remember this mnemonic: the seven layers. Please do not throw the sausage pizza away. That’s right. Sausage pizza is so yummy and so delicious that we wouldn’t dare throw away a piece of sausage pizza.
So please do not throw sausage pizza away. If you take the first letter of each of those words, it will match up with the seven layers of the OSI model and help you remember it come test day. Now, the next thing we want to talk about in the OSI model is the way data flows through the network. And as the data goes through different layers, we call it different things. So when you interact with data, you’re usually operating at layers 5, 6, and 7, the session, presentation, and application layers. And so we call it data. As we move lower through the layers, we’re going to go down into segments, packets, frames, and bits. And so this chart on the screen is one you’re going to want to memorise as well. Now, don’t worry, I’ve got a pneumonic for this one, too. And that is, don’t some people fear birthdays? And we all get afraid of getting a little older, right? Well, this is going to help us remember that. Don’t let people forget that birthdays are data. segments, packets, frames, and bits.
Layer one is physical. So the first layer we’re going to dig into is the bottom layer, or the physical layer, of the OSI model. In this layer, bits are transmitted across the network, and this is going to be focused on the physical and electrical characteristics of the network. So, for example, if I use an Ethernet cable, such as a cat-5 cable, that is a wired network that runs over copper, or if I use a MTRJ fibre optic connector on a fibre cable, that is a physical choice. Or if you’re using wireless networks, you’re using radio frequency; all of those have physical and electrical characteristics. But regardless of the method that I use when I’m setting data over it, at layer 1, I’m using bits—ones and zeros.
That’s it. So the characteristics that we’re going to define in the physical layer are how the bits are represented on the medium, which is our wires, our cables, or our radio frequency, and the wiring standards for the connectors and jacks. Am I going to be using an RJ-45 or an RJ-eleven, depending on if I’m using a phone line or a network? We’re also going to focus on the wiring standards for our connectors and jacks. Am I going to use an ST connector, an SC connector, or an Metro connector? And those are all fibre connectors we’ll talk about later on in this course. Or am I going to use an RJ-45, which is a Cat-5 connector on a copper cable? We’re going to also talk about the physical topology. So when we talked about topologies, when we talked about the hub and spoke, the bus, the ring, and the star, the way we have that physical topology laid out is layer one. We’re also going to talk about synchronising bits, bandwidth utilization, and multiplexing strategies. We’ll get to those in a few minutes here.
So how are bits represented on the median? Is it an electrical voltage with copper wiring, or is it light? If you’re using fibre optics, each of these is going to be used in a different way to represent a one and a zero. For example, if I’m using a copper cable like a Cat5 cable, I might use 0 volts for a zero. And if I have a plus five or a negative five, that can represent a one. And so this switching between the zero and the plus or minus five is going to tell me if it’s a one or a zero on the network. And we do this through a method called transition modulation. If it’s changed during the clock cycle, then a one is represented and if it’s not, it becomes a zero. Now, do you need to know this specific detail for the Network+ exam? No, they’re not going to ask you about transition modulation. It’s just too deep in the weeds. But I want you to understand the concept here. Now, if we’re dealing with light, we can do the same thing. But instead of voltage, we’re just going to turn a light on or off. And if we leave it on or off, those transitions can be used to determine whether it’s on or off via transition modulation. Now, the way your cables are wired is our next thing, and that’s our next characteristic.
When we use an RJ-45 cable, we adhere to the TIA-568-B standard. That is the standard wiring, if you remember the 5/6/8B part. Now, crossover cables will use an A on one side and a B on the other, and what that means is that we’re actually going to flip the transmit and receive pins on the end of the cable, and that’s why the A and B are different. Straight through cables, though, they’re going to use B on both sides. Although some networks have been wired with A on both sides, the standard is 568B on both sides of a straight through cable. Now, in the Ethernet module, we’ll go over wiring standards in great detail. In fact, we’re going to talk about every single one of the eight pins, what they do, and how to wire them up. And you will have to know that for the exam because I have seen test questions in the past asking you to cable up a wire in simulations or in multiple choice questions. So keep that in mind as we’re going through this material. Now, how are your cables connected? This is the physical topology piece. Layer 1 devices view networks from a physical topology perspective, and this is going to tell you whether it’s a bus, a ring, a star, a hub and spoke, a full mesh, a partial mesh, or any other topology that you want to be using. The next thing we’re going to talk about is: how are you synchronizing this communication? So if I’m sending ones and zeros to you, are you ready to receive them? Well, there are two ways that we could be talking.
We can do it asynchronously or synchronously. Now we are currently doing this asynchronously. I’m recording this lecture, and you’re going to watch it weeks, months, or maybe even a year after I record it. And that’s asynchronous communication, right? We do it out of sync and out of time. Now, when we do it in networks, we use what’s called a start and a stop bit. It’s kind of the same thing you do here on the video. When you’re ready to receive the information, you hit the play button, and it starts sending it to you. When you don’t want it anymore, you hit the stop button, and it stops sending it to you. Well, with networks, we send a start bit that says, “I’m ready to start talking,” and then it starts talking, and when it’s done, it says, “I’m done talking” by sending a stop bit. Now, if we do things synchronously, we have to be in the same place at the same time.
And so for our networks, we would use the same clock. For example, if I was running this in an in person classroom, we might say that at 08:00 A.m. On Monday, we’re going to start class. So I would look at the clock in the classroom, and when it turned 8:00 a.m., I would start talking. If you’re not ready, you’re going to miss the information. Right, that’s the idea of synchrony. We’re both using the same time to do the communication. Now how are we going to utilize bandwidth? Well, there are two main methods that we use. One is called broadband, and the other is called baseband. Broadband is going to divide your bandwidth into separate channels. The easiest example of this is cable TV. If you have cable TV in your home, you may have 200 channels coming in, but you only have one cable. That’s bringing all of that content onto your TV. You then choose, based on the channel, which part of that you want to listen to, and you filter out the rest. Baseband, on the other hand, will use all of the frequencies all of the time over a single cable. So, for example, in your telephone, you use baseband communications.
When you pick up the phone, you’re using all of the bandwidth that the phone company has allocated you for that phone line. You can only make one phone call at a time because you’re using the entire bandwidth. You’ll use a reference clock to accomplish this, which will allow you to send information that both the sender and the receiver can access. A good example of this is the wired networks in your house. If you use Ethernet in your office or home, that is going to be a baseband network. So how can I get more out of a limited network? So if I’m using baseband, like a copper network or an Ethernet network like we just discussed, and I’m using all of the bandwidth, how can I get more out of it? Well, we have a couple of mechanisms here, and they all have to do with division multiplexing. The first one is time division, where each session is going to take a turn using dedicated time slots, and we’re going to share that cable between all of the users on that network.
So if you want to think about it, say you have to use the bathroom in your house, and you only have one bathroom but four family members, you can’t all use the bathroom at the same time. So maybe you’re going to use it first, and then your sister is going to use it, and then your brother’s going to use it, and then your mom’s going to use it. You just take turns. You each have your own time to use it. The next one we have is statistical time division, multiplexing, or stat TDMA. And this is a more efficient version of time division multiplexing because it dynamically allocates these time slots on an as-needed basis. So, for example, in the bathroom example I used, we really were talking about STDM there. If I were talking about pure TDM, I would actually allocate you a time frame. So you get to go to the bathroom first from eight to nine, then your sister from nine to ten, and your brother from ten to eleven. The problem with that is that if your sister didn’t have to go to the bathroom from nine to ten, well, nobody could use the bathroom because that was her allocated time. With Stat TDMA, you look at it and go, “She’s not using it; I’m going to jump in there, right?” That’s the idea.
I know it’s kind of a silly example, but hopefully it’s fixing your head. Now, Frequency Division Multiplexing, or FDM, is going to take the median and actually split it up into channels, similar to the way broadband works. And that’s exactly how broadband works, right? We take this single cable and break it up into 5100 or 200 different frequencies that can then be used, and each person gets a piece of that to be able to use it. That’s the idea behind FDM. Now, for the exam, do you need to memorise TDM, stat TDM, and FDM? No, you don’t. It’s just something that, if you see the word “multiplexing,” think about the fact that we’re trying to share a limited resource in a more effective way, but they’re not going to ask you which of these and to identify examples on the exam. Now, in the real world, you may come across some of this, but for the exam, don’t worry too much about it. Now, what things operate at layer one? Well, pretty much everything. But specifically, when we talk about what is a layer-one device, we’re talking about things like cables, Ethernet, and fibre optics, for instance.
We also talk about radiofrequencies, Bluetooth, and WiFi. We also talk about infrastructure devices like hubs, access points, and media converters. So if I asked you any of these things on the screen, what layer they are, they’re all at layer one, the physical layer. They operate with bits, and all they do is repeat whatever they get in. So if something goes into one end of the cable, it’s going to come out of the other end of the cable. There’s no routing; there’s no switching. same thing with a hub. Whatever comes in on one side goes right out on the other. That’s the idea of layer one. Stupid devices that simply
Layer two is the data link layer. So the layer two datalink layer is going to package your data into frames and transmit those frames on the network, performing error detection and correction, uniquely identifying the network devices using a Mac address, and providing flow control. Now, what is a Mac address? Well, it’s a way to do physical addressing and operating rates on the logical topology. So where the physical devices were concerned—how things were physically cabled—now we’re going to think about how they’re logically connected together. And this is going to be the method of transmission used at layer two. When we get to layer three, we’ll use a different method. Now, LLC, or Link Layer Control, is used for flow control. This will perform the synchronization of your transmissions and connection services. We’ll talk more about that in this lecture.
Media Access Control, or Mac address. Mac addresses are really, really important when you’re dealing with switches and layer-2 devices. Now, this physical addressing system uses 48 bits assigned to every single network interface card on the planet, and it’s assigned by the manufacturer. If you see it in red, it is a twelve-digit hexadecimal number. That is what a Mac address looks like. Because they’re written in hexadecimal, each of those numbers or letters is actually considered four bits. So the first 24 bits, or the first six characters, identify the vendor. The last six in my example here, D 251, F 1, are going to identify a unique value in the United States. It works much like a Social Security number does for a person. You can tell where a person was born, what state they were born in, and roughly what year they were born by looking at the first three digits of their Social Security number. and if you look at the rest of it, it’s going to uniquely identify that person. So if my social security number was 123-45-6789, the second half of that is going to say, “This is Jason,” and the first half is going to say he was born in X state.
Now, with a Mac address, that first half is going to tell you whether it’s an Apple device, a Dell device, or a Raw Link device. The second half will reveal who exactly owns your device. Now, we use this in the logical topology, and the layer 2 devices are going to view networks logically, whether they’re ring, bus, star, mesh, hub and spoke, or whatever else they are. But it’s all based on this Mac address of who talks to each other and what it logically looks like. The method of transmission we’re going to use here is interconnecting all of these devices together over some physical mechanism. And we don’t care about the physical mechanism. That’s a layer one problem, whether it’s wireless, fiber, or copper. But we do care about determining whose turn it is to talk and transmit to make sure that we’re not talking over each other. The best example I can give you of this is when we’re sitting in a classroom and I have 20 students. There’s only one of me, but there are 20 of them. If all 20 of them started shouting out things, I wouldn’t be able to understand who was asking the question. So instead, we have a mechanism for figuring out who’s going to go first.
And in the classroom, we use this by raising our hand, right? You raise your hand, and then I go, “Oh, Johnny, what’s your question?” Then he explains to me how we control information in a network. We use electronic mechanisms to do that, but it’s the same kind of concept. The next piece we have is logical link control, and that’s LLC. Logical Link controls are going to provide connection services. It’s going to allow for an acknowledgement of receipt of the messages. So if I told you something afterwards, I could say, “Did you get it?” And you could say, “Yes, I get it.” You acknowledge the receipt, and I can move on. That is what LLC does in our networks. It is the most basic form of flow control. It’s going to limit the amount of data that a sender will send at one time to keep the receiver from being overwhelmed.
If I talk too fast for you, as you’re trying to jot down notes, you have no way to tell me. So you would just drop things on the floor, and we would keep on moving on. In a real classroom environment, you might raise your hand and say, “Hey teacher, you’re going too fast; can you slow down?” Or I don’t understand that concept; wait a minute. But in a virtual environment like this, we don’t have that ability. Well, that’s the idea behind logical link control. It allows the device to go: “Hold on, you sent me too much; I’m not ready to receive anymore.” It’s also going to give you some basic error-control functions. And this is going to allow the receiver to let the sender know that the expected data frame that they got wasn’t received or was corrupted. This is accomplished through the use of a checksum. So essentially everything it’s getting is a series of ones and zeros. And so it can add up all the ones and zeros, and the last bit will either be even or odd. And if it matches, they assume that it was good; if it wasn’t, they figured it was bad, so they asked for a retransmission.
So how is communication synchronized? Well, in addition to using the asynchronous and synchronous modes that we had back in layer one, we’re going to add a new mode called isochronise. And ISO stands for the same thing. Cronus means time. And so what this really is when network devices are going to use a common reference clock source, much like we did with synchronous, but the difference is we’re going to create time slots for transmissions. And so this is going to have less overhead than synchronous or asynchronous communication because the devices are going to know when they can communicate and for how long. The next one we have is synchronous, much like we had back in layer one. When we talk about synchronous, we mean that the devices use the same clocking mechanism. And they’re going to do this to indicate the beginning and end of the frames.
And they’re going to use control characters or separate timing channels to keep that synchronicity going on. So if you’re a music person and you hear a song on the radio, you can usually pick out the beat, right? And it could be done in three, three, four, or four, four times. So, for instance, I listen to a lot of music that has four-four timing. Now what does that mean? It means it’s going to beat every measure four times every measure. And so, if we were using that as our communication mechanism, to give an analogy, I would only be able to talk on the beat. So I am talking on the beat. See how I have to wait? And there is a lot of downtime that is underutilized. That’s one of the functions. The disadvantage of asynchronous synchronicity is that each network device references their own clock cycle, and then they use those start and stop bits that we discussed last time. So what are some examples of layer two devices?
Well, network interface cards, as you see here on the screen, Additionally, we have bridges and switches. Now, we talked about hubs being really dumb, or whatever comes in, they just send it right back out with switches. They are actually smarter. Using those Mac addresses, they can begin to learn logically which physical ports on the switch those interfaces are connected to. And so, for instance, if I know in a classroom that I have 20 students and I can look at the desks and say, “Okay, Johnny is in desk number one, Mary is in desk number two, and Susie is in desk number three,” When Susie wants to pass a message to Johnny, she can give it to me, and I can put it on desk one because I know where everybody is sitting. It’s the same thing with a switch. When your network devices talk to the switch, it learns their Mac address and what switch port they’re on. And so it starts sending data only to those devices at the location where they’re residing.
Three network. So when we get to the network layer, we start getting concerned with routing. And this is all about how we’re going to forward traffic, which we refer to as “routing with logical addresses.” For example, your computer has an IP address, and that IP address is either going to be IP version 4 or IP version 6. Both of those are layer 3 protocols, and we’ll talk more about them as we go through.
The other thing we’re going to be concerned with is logical addressing. Now, I mentioned IPV-4 and IPV-6, but they are not the only logical addressing schemes out there; they’re just the most common. We’re also going to be concerned with switching, and we’re talking about layer-three switching. We are actually referring to routing, but it is called “switching” at layer three, which can get confusing considering switches are layer-two devices. And then we have route discovery and selection. How do I know which way I want to go? We’ll talk about that a little bit more here too. We’ll talk about connection services, bandwidth utilization, and multiplexing strategies.
All of this is at layer three of the network layer. So logical addresses There are numerous routing protocols that have been used over the years. There was Apple Talk for Apple computers. Inner Network Packet Exchange (IPX) was available for Novell Network computers. And even Windows computers use that. But then the Internet Protocol came out, and it kind of killed all the other protocols. Some of those are still on legacy systems. But the routing protocol of the Internet and of your home network is going to be IP. IP comes in two variants. IPV 4 and IPV 6 are available. If you look on the screen here, this is an example of an IP address: 170, 216, and 254 one.
And we’ll look more at IP addresses in a separate lecture as we dig deeper into routing later. How should data be forwarded or routed? Well, there are three main ways of doing it. You can use packet switching, circuit switching, or message switching. The most commonly used one in your networks is going to be routing or packet switching. This is where data is divided into packets and then forwarded on. I like to think of this like an envelope that I would put in the mail. I’m going to write the address of my mom and the city and state that it’s going to go to on it. I’ll put it in my mailbox, and the mail carrier will take it to a central location.
From there, they’re going to figure out what state it belongs to and send it to their local post office. And then once it’s in that state, it’ll get down into that city, and then from that city to that street address, and then eventually to her address. And it’s going to keep switching that packet, that one envelope, until it gets to the final destination. That’s the way packet switching is going to work. Now, every time I send her a letter, it can take a different route. I don’t really care as long as she gets that letter. It’s the same thing with our packets in our network. Now, when I talk about circuit switching, this is where we get a dedicated communications link that’s established between two devices. So if I pick up the phone to make a phone call, I’m going to actually make a virtual connection from my phone to the end receiver’s phone.
And our voices will continue to follow the same path throughout the conversation. Now, when we hang it up and make another phone call, it may take a different path that time, but for the entire session, it’s going to use one path. That is the way circuit switching works. The final one is message switching, and message switching is where data is divided into messages, similar to packet switching. But these messages can be stored and forwarded. So if you go back to my mail example, maybe it gets to my mom’s state and the post office is closed for the night. Instead of just dropping that packet on the floor, it would hold it there until they open and then push it on along the way. The same thing happens virtually with our networks if we’re using a message-switching protocol. But almost all of our networks nowadays and the ones you utilise are going to be packet switched, unless you’re dealing with backend networks, where you may see some circuit switching or message switching.
Now, route discovery and selection How do we decide which path we’re going to take when we send a message? Well, routers maintain a routing table to understand how they can forward a packet based on the destination IP it wants to go to. There are lots of different ways they can do this, and they can do it either as a static route or a dynamically assigned route using a routing protocol like Rip OSPF or EIGRP. Now, we’ll talk about those and many other routing protocols when we get into routing and talk specifically about how this works. But for now, let’s just use the example on the screen to give a really basic idea of how routing works. Assume I’m at router number five in the bottom right corner and want to get to router number one. How should I do it? Well, I can go from five to four to one, and that would work, but I could also go from five to four to three to two to one, and that would work as well. How do I know which way is the best way to go?
So, if I’m using a dynamic protocol, all those routers are communicating with one another to determine which route is the fastest. If you think about this like a street, when you type in your GPS and you’re trying to go to the grocery store, It may take you three different ways, depending on the time of day, the traffic, and the congestion. Routers do the exact same thing. They all talk to each other and say, “Hey, I’ve got a better way for you to get there because there’s too much traffic over here. You should take this other route.” That’s the idea with route discovery and selection. The next one we have is Connection Services, and Connection Services are going to augment the layer 2 connection services and provide additional reliability. Again, we’re going to have some more flow control, and this is going to prevent the sender from sending data faster than the receiver can get it. So again, there’s that flow control to say, “Hey, slow down; you’re sending me too much,” or “Speed up; I can take more.”
We also have packet reordering. Now packet reordering is really important because it allows us to take this big chunk of data and cut it up into little packets and then we send those packets off to different directions until they finally get to their destination. Now, the problem is that sometimes they’re going to get there out of order. And so packet reordering allows them to get all that data at the end when it gets to the receiver, and he can go and say, “Okay, I got packet one and five and two and four and three.” And then he’ll put him in order 12345, reassemble it, and he’ll have all of his data in one piece. That’s the benefit of routing: that each of these packets gets numbered and sequenced. So when they get it to the other end, even if the pieces are out of order, they can be put back together into a coherent message. The next thing we’re going to talk about here is the Internet Control Message Protocol, or ICMP. ICMP is used to send error messages and operational information about an IP destination, and the most commonly used one is Ping Pin G, which we’ll talk about specifically in the troubleshooting lecture. But as you can see in this example, we can send out a packet, which is a test packet, and when it comes back, we can then say if the site is up or down. In this case, it was because we got five responses back.
It’s not regularly used by end-user applications, but it is used by us as administrators to troubleshoot our networks. And again, the most commonly used one is Ping and a variation of that called Trace Route, which will trace the route a packet takes throughout the network and tell you every single router along the way as it does. So what are some examples of layer-three devices? Well, we have routers and multilayer switches. The icon here that you see on the screen with the four arrows is a depiction of a router in a logical diagram. We also have the IP Version 4 and IP Version 6 protocols, and we have ICMP, which is the Internet Control Message Protocol that we just talked about and that we use in troubleshooting. All of these are found at layer 3. The best one to remember is IP and routers, because those are going to be the common ones you’re going to see on test day.