 Hi and welcome to the video courses for Windows Server 2016. My name is Patrick Loner and I'll be your instructor. Let's start with a little bit about my background. I've been in the IT industry for just about 18 years. I got my start with an MCSE on Windows NT and have sort of worked my way up through the ranks on every version of Windows that has existed. I began my career in a position of network administration at a training center where I handled just about everything. Imaging of PCs, hardware and software, troubleshooting in classroom environments, as well as for the business side of things. Touched just about everything and sort of felt like a jack-of-all-trades master of none. I quickly got into training, some of the Windows 2000 courses and CompTIA courses and as they say, the rest is history. And I've been working as a Microsoft certified trainer now for the better part of 18 years. I've had a couple of positions, one with the training center, another with a network consulting firm for three years where we were involved heavily in projects for upgrading to newer versions of Windows Server, upgrading and migrating to exchange. For the past 10 years, I've operated as a freelance trainer and network consultant. It's my pleasure to be your instructor on these courses and let's get ready to get into the material. In the first topic, we're going to be dealing with planning and implementing IP addressing schemes for IPv4 networks. As I said, we're going to be dealing with some of the fundamental, not basic, but fundamental steps in configuring IP addresses. IP addresses are gonna be required for every host on the network and it's not as if we just assign an IP address off the top of our head. No, the IP addresses have to fit a particular structure. Each individual host on a single network segment has to have an IP address that works with the other host. It has to have an IP address that works with the router, which is their default gateway. There's a rhyme and a reason to why we assign these addresses. And some of us are gonna find ourselves in networks where we have multiple subnets and we have to take a single address and make network IDs for the multiple subnets and then assign individual host IDs in those subnets. Now, even if you don't find yourself in that situation in real life, you're going to find yourself in that situation on the exam. Microsoft will test your knowledge at every different part of the subnetting process. And so it's incredibly important that we're able to answer these questions and we are able to design IP address schemes for IPv4 networks. So let's start just by talking about the general networking requirements. What is required in order to connect a system to the network? You know, most of us get spoiled. We just plug in a network cable and we're off and running. We're connected to the network or we are sitting at the coffee shop and we activate our wifi and we check a box that says we accept some agreements and we're off and running. We're on the internet. And the vast majority of society has absolutely no idea what's going on behind the scenes. Well, that can't be you. And you know that can't be you. That's why you're here. You need to understand what's happening behind the scenes. And so the three components that are required in order for a system to connect to the network are the NIC, the protocol and the client, all right? Now, why do we connect to networks? We could cover that initially. We connect to networks because we have to or because we wanna share resources, right? The purpose of a computer network is to connect systems together to share resources, files, folders, printers and internet connection, an app. Those are the reasons that we network computers. And we've been networking computers for many, many years and it's always required these three components. The network interface card is the first. This is the physical interface between the system and the rest of the network. Now the network interface card maps to a particular network architecture. So it connects you to a very specific kind of network. Now today, this has been simplified. You know, your network interface cards are going to either be ethernet or wifi. It's very unlikely for any of us to encounter a token ring card or an arc net card. Something like that. These are older and equated technologies that just simply aren't even used today. So ethernet is the predominant standard on today's local area networks, providing varying speeds anywhere from 10 megabits per second all the way up to really 10 and 40 gigabits per second. But the predominant speed on local area networks for clients and servers would probably be the one gig per second gigabit ethernet. This is a wired internet connection or a wired network connection, I should say. And then we have wifi standardized as 802.11 and various types of that exist as well. And that's just using radio frequencies instead of copper cabling to communicate. But in any case, and we're not quite getting down to that level, in any case, I have to have a network interface card. And it is going to physically interface with the network. But beyond that, we have to have a network protocol. And a network protocol functions very much like a language. So if you and I are, you're listening to me and I'm speaking English and you understand English, then we can communicate with one another. However, if the only language you spoke was French and the only language I spoke was English, then we'd have a problem. We wouldn't be able to communicate. And that's the way computer systems are. They have to have a common network protocol in order to be able to communicate with one another. Going back to that analogy, I could speak four languages and you could speak one. So long as one of the languages that we speak is the same, then we can find some common ground. And now a network protocol is more than just a language. It defines the rules and techniques for communication. But that is the best analogy. Now thankfully today, we've standardized pretty much across the board. TCPIP, Transmission Control Protocol, Internet Protocol is the standard protocol on the internet. And it's the standard protocol on nearly all local area networks and with all major network operating systems. TCPIP is a routable network protocol that uses IP addresses to identify individual nodes. So that's one of the parts of the protocol. A unique address for the individual systems so that they can communicate with one another. And we'll get far more into those addresses here in just a bit. The third component that you need is a network client. The two network clients that are installed by default on Windows systems are the client for Microsoft networks and then file and print sharing for Microsoft networks. The first one allows you to log in to a network and communicate with other systems. And the second allows you to make your resources available on that network. So these are the three components that you require in order to connect any system to a network. Now let's get into the details of the protocol itself. TCPIP, which right now we're just referring to as IPv4, there is an IPv6, but we'll cover that later. The current standard that's used on most networks is IPv4. So when you use TCPIP, you have another three components that are required by the protocol itself in order to provide network communication capabilities. The first is an IP address. It says an IP address that functions very similarly to a mailing address in that it's unique to that particular node. That term node just simply defines a system on a TCPIP network with an address. Okay, so just like if I want to receive mail through the postal system, I have to have a unique address that's the same on a TCPIP network. Without a unique address, then you can't communicate. You also need a subnet mask. And the subnet mask is gonna be used to differentiate between the network and the host portions of the address. Now we'll talk about this more as we go along, but you can kind of consider this like the house number and the street name of the IP address or of your physical address, okay? So if I live at 123 Oak Street, for instance, and the adjacent street is 123 Maple or is Maple Street. Well, there's another house, right? That has the house number of 123 on Maple Street, but the postal worker doesn't have any problem delivering mail because it's going, there's two parts to the address. There's the street name and then there's the house number and that's the same way that a network address is there's a network ID, okay, the street name or could be called the street name, I should say. And then there is the house number, which is your individual host ID, right? It's the subnet mask that's the component that is used to differentiate between those two parts of the address. Then you have a default gateway and the default gateway is your ability to route packets destined for remote network segments. So technically are you required to have a default gateway? Well, no, I mean, in reality, you need an IP address and a subnet mask. You don't have to have a default gateway, but the vast majority of us will want to communicate with the internet in many network environments and even in local area networks, you'll have multiple subnets. So you'll have a system that has to communicate with systems off of its local network segment, right? In the postal system, I want to send mail to my people potentially on my own street, but I also want to send the mail to people in other cities and maybe even in other countries. The postal system, my default gateway is my mailbox because the postal worker is going to come and pick up the mail. In the TCPIP network, my default gateway is my router. And so I send packets to the router and then the router is going to worry about where those packets go beyond that to get them to their ultimate destination, all right? So just like in the regular mail system, in TCPIP, I need that way to get packets off of my local network segment, and that's going to be the default gateway. All right, so let's focus in on this IPv4 address because this is incredibly important. IPv4 addresses are 32-bit binary numbers. The binary numbering system, as we will discuss, and as you may already be aware, is a base two numbering system, okay? That means it only has two values. So decimal as a numbering system has 10 values. Those values are zero through nine. In the binary numbering system, you only have two, zero or one, okay? So a 32-bit binary number would look like what we see here. Now, thankfully, because we're humans, the developers of TCPIP and the various operating systems decided that that would be a little bit cumbersome, making us write out 32 zeros and ones. We thank them for that. So we get to write IPv4 addresses in what's called dotted decimal notation. So it's a 32-bit number, but it's divided into four groups of eight bits, or bytes, and those bytes are separated by periods. So we get numbers like 192.168.1.200, 172.16.123.100. So on and so forth, all right? But each of those digits that are separated by the periods represent a byte, eight binary bits. And that's important because we will understand, if you don't already understand, that a byte only has certain possible values. So the minimum value of a byte is all zeros, which equates to the decimal number zero. The maximum value of a byte is all ones, and that equates to the decimal number of 255. All right, so you can never have a number in an IP address that is greater than 255. So 127.0.0.1 would be a valid IP address. 127.255.256.255 would be an invalid address. And it's just because it includes a number that is not possible in eight binary bits, all right? So 32-bit numbers written in dotted decimal notation. We, that division, and I didn't put this on the slide, but that division is referred to as an octet, okay? So you'll hear me say in the first octet, in the second octet, and so on and so forth. And so when we refer to octet, we're just talking about that grouping of eight bits. The first octet would be the first group, the second octet, the second group, so on and so forth. Now when we start digging in, we need to understand that the IPv4 address is actually comprised of two parts. And we've already discussed this a bit, the network ID and the host ID, similar to the house number and street name. So the host ID would be like the house number, the network ID would be like the street name. The network ID identifies the network segment that the node resides upon. The host ID identifies a unique identifier of that host on that particular segment. So just like my house number identifies me uniquely on my street, so does my host ID on my network ID. In other words, if you had two networks, 192.168.1.0, a second network of 192.168.2.0, we could have hosts that had the host ID of 100 on both of those networks, and that wouldn't be a conflict, just like 123 Oak Street and 123 Maple Street don't cause a conflict because the street name is different, right? So the host ID has to be unique within the network ID and then collectively that makes the entire number unique. So it does mean though that we need another component. In street addresses, we know there's the number and then there's the name, but in the IP address, it's all part of one 32-bit number. So how do we tell the difference? How do we know where the network ID stops and the host ID starts? We know based on the subnet mask. So the subnet mask is another 32-bit binary number, also written in dotted decimal notation that differentiates between the network and the host ID portions of the address. The subnet mask is a little different though in that it consists of continuous ones followed by continuous zeros. So the subnet mask always starts out with ones, but once the ones stop and the zeros start, the ones never come back, and we'll discuss this a little bit later, but in an IP address, you can have any valid value from zero to 255 because those are the values that you have in an eight-bit number. That's not the case in a subnet mask. In a subnet mask, there are only certain possible legitimate values, 128, 192, 224, 240, 248, 252, 254, and 255. And that is because those are the values that we get with continuous ones. To get any values other than that, we would have to have things like 111, 00, 111. You'd have to start stopping and starting the ones. And you can't do that in a subnet mask. So continuous ones followed by continuous zeros, essentially when the ones stop, that is when the mask stops and that is going to be pointing out the network portion of the address, but we'll get into that more when we can see it a little bit better. Now these can be expressed and written in dotted decimal notation, and they typically are, especially in the Windows operating system. They can also, though, be expressed in CIDR notation. CIDR is classless inner domain routing, and it is a mechanism through which we can specify the number of bits in the subnet mask that are set to one. So you might see an address listed as 192, 168, 1.100, slash 24, and that 24 represents the number of bits out of 32 in the mask that are set to one. So it's just another way of writing out the subnet mask. It's actually an easier, a simpler way. The third component that's required is gonna be the default gateway. As we said, this isn't necessary for general communication, but it's necessary to communicate off the local network segment. Your typical network clients are connected to a single network segment, and they're only able to communicate with their own subnet. So if they try to communicate off that local subnet, they have no idea how to get the packets to that location. So they require a network router. A network router is a device that is connected to multiple segments. Now the network router is not connected to every segment in the world, of course, but routers are strung together through the use of ISPs to get packets to their ultimate destination. Couple keys about the default gateway, which are really important for us to understand is that if the default gateway is my way off my local network segment, and I don't have the ability as a network host to communicate with remote segments without a router, then it would stand to reason that the default gateway has to have an IP address on the same network segment, aka subnet for proper communication. This is gonna be really key when we start troubleshooting because we're gonna look for the presence of a default gateway on a client that's, say, having trouble communicating with remote networks, but not only do we need the gateway to be present, but we need it to be accurate. So if my network is 192.168.1 and my gateway is on the 192.168.2 network, then I'm not going to be able to communicate. I can't, you know, and I think you probably get the idea. So it's gotta be on the local subnet, and it's often the first or last address in the range, 192.168.1.1 or 1.254, but the gateway is actually a router, or, better said, it's an address that's assigned to a network interface on a router, but it's configured on the clients, and the clients know that when they have any network packets that are destined for IP addresses that reside on a remote subnet, they must forward those packets to their default gateway. Now, originally when IPv4 was introduced, the Internet Assign Names and Numbers Association, IANA, organized IPv4 addresses into classes, and this is referred to as the Classful System. Each of these classes would have a default subnet mask. Now, this is no longer used technically. It was a very wasteful system, and so we've switched to classless routing, hence the old CIDR or Classless Interdomain Routing syntax, okay, but it's consistently referenced. So, you know, when we get in, we still have to deal with it because when we get into subnetting, you know, we'll say things like, okay, we'll start out with a Class C address and then create subnets, or start out with a Class A address and then create subnets. Well, there are no Class A addresses anymore, but we, again, the terminology is still out there, so it's important for you to understand, and it's not just a historical thing, it's just the way that people, network admins still communicate, okay? So, the class of the address was defined by the first octet. The value in the first octet. And then the class defined the default mask. Well, because the subnet mask differentiates between the network ID and the host ID, it determines how many bits are available for network IDs and how many bits are available for host IDs. And the more bits you have available, the more potential IDs, right? And so that's where these classes come into play. So Class A, the first octet would be between zero, or excuse me, one and 127. And the default mask is an eight-bit mask, 255-000. That means you only have the first eight bits for network IDs. So it's two to the eighth, which would technically be 256. However, we've got some reserve characters in there and we won't get into all of this, so let's just take my word for it. I mean, basically you can't modify the first bit, all right? So it's technically two to the seventh, and it's minus two because you can't use all zeros and all ones. So 126, 126 possible class A networks. Not very many, right? This is in the entire world based on the IPv4 address space. However, each class A network could support over 16 million host. Okay, how many organizations do you think that would need 16 million host? Well, not that many. ISPs would be about the only one. And so class A networks, few and far between, as to the actual organizations that would need that size. This doesn't mean these can't be subdivided, so we'll get into that later. That's called subnetting. Class B, 128 to 191 in the first octet, we need to recognize that as a class B address. It's got a 16-bit default mask, 255, 255, 00. A possibility of 16,384 networks. And on each of those networks, we could have up to 65,534 host. Class C, the first octet is 192 to 223. Default mask of 24 bits, 255, 255, 255, 0. So a little over two million of these networks, but on each network, you could have a maximum number of 254 host, okay? So as I said, the classful system's not used anymore because it was wasteful. We can really quickly see how it was wasteful. IPv4 was originally introduced with the goal of assigning public IP addresses to every system that was connected to the internet. So an organization would call up an ISP or a registrar and want a public address space. And they would say something like, okay, well, we've got 650 systems, okay? And we look at this table and we say, well, the only way that we can give you 650 public addresses would be either to give you three class C's or a class B. Well, the problem with giving you three class C addresses is it clogs the internet routing tables. These are the tables that routers use to make decisions as to where to forward packets. So you'd have three IDs all pointed at the same place and it causes clogging. The alternative though would be to give you a class B address when you only needed 650 addresses, you would waste nearly 65,000 addresses, okay? So the classful system works on those decimal points in the subnet mask. It goes from eight bits to 16 bits to 24 bits. And so therefore it was very wasteful. That's why we're not using it anymore. Having said that though, and as I said before, we've got to be able to look at the first octet, recognize the class, recognize what the default mask would be. That's gonna be important from a subnetting perspective. Now we've mentioned subnets a few times, exactly what is a subnet? Well, a subnet or subnetwork is simply a smaller network segment on a larger TCP IP LAN. And each subnet is uniquely identified by a network ID. You can use the term network segment if you want, it really means the same thing. Within a subnet, all the hosts are gonna share the same network ID and they're all gonna share the same default gateway because the default gateway is gonna be the IP address of the router that's connected to that subnet, all right? So at least one router on the LAN has an interface that's assigned an address for the subnet, it's physically connected. It's the way for those hosts to get off of that subnet. Essentially what some organizations are gonna have to do is begin with a single classful IP address and then create subnet ranges. Now some organizations that you find yourself dealing with will just have a single network segment. Okay, so I have a small organization maybe has 50 clients, for instance, where I started. We just had one subnet. There wasn't a need for multiple subnets. We had an internet connection, it came into a router, that router had one interface, it interfaced with network switches, all the clients connected to the network switches and that was it. Everybody was on the same subnet. There was, because we just didn't have that many systems but some organizations are going to have multiple physical segments, multiple router interfaces and each of those interfaces and segments has to have a unique ID. So what we typically do is we start with one address and we break it into multiple subnet IDs and that is the concept of subnetting. So when we talk about creating subnets, we're referring to the subnetting method and that is the method that an organization is going to use to take a single network ID and then divide it into multiple network IDs. And this can be a confusing process, okay? And it's confusing in some ways just because of terminology and other ways it's confusing just because we have to use binary math. Now we've just seen binary computation and understanding how the zeros and ones equate to decimal numbers. We know that the subnet mask differentiates between the network ID portion and the host ID portion of an IPv4 address. So the process of creating subnets is consisting of using bits that were formally associated with the host ID and now using them as a network ID, okay? And we do that in order to create a unique what's called a subnet ID. So we're technically adding onto the original network ID and we're now including more bits. Well, those individual bits that are included now are called the subnet ID, all right? So let's just look at an example, for instance, a 10.000 slash eight network uses eight bits for the network ID and then that leaves 24 bits out of the 32 for the host ID. So the process of subnetting would borrow bits from the host portion to create a subnet ID. And so let's just say we borrow three bits, all right? So we, in other words, when we say borrowing bits we're talking about extending the subnet mask three bits. So now it's 10.000 slash 11, okay? Now that gives IDs for up to eight subnets and it still leaves 21 subnets for the, or 21 bits, excuse me, for the host ID, okay? So we're 11 and 21 equals 32, right? Now you might be wondering how does that give IDs for up to eight subnets? We will discuss that. There's a mathematical formula for determining the number of subnets that have been created and it's all related to the number of possible combinations of zeros and ones. We've added three bits and in those three bits there are eight possible combinations of zeros and ones. So we've provided three additional bits for the network ID and so that creates the subnet ID, all right? The method of changing the purpose of the bits is actually performed by manipulating the subnet mask. We are extending the subnet mask and in extending the subnet mask we're saying, you know what, the subnet mask first said that only the first eight bits were a part of the network. Now it's saying that the first 11 bits are part of the network ID or reference the network ID and thus we give ourselves additional bits to play with to create these unique subnet IDs. Now let's walk through the actual process for subnet creation and then of course we're gonna take a look at this because without taking a look at it it's mumbo jumbo for lack of a better term, all right? So the process begins with starting with a classful network ID and it's subnet mask. Like we just did, a 10-0-0-0 network is a class A network and its default subnet mask is an eight bit subnet mask. The second step is to identify the necessary number of subnets as well as the appropriate number of hosts per subnet. Okay, so we gotta make sure that we create the correct number of subnets but we also leave enough host bits to still accommodate the number of individual IDs that are necessary on each network segment. Once we have those two figures then we can extend the subnet mask and we're extending the subnet mask using a formula two to the power of n. Now the letter there doesn't matter but it's two to the power of n where n equals the number of bits that have been added to the subnet mask. So using our previous example we went from an eight bit mass to an 11 bit mass so we added three bits. So two to the power of three equals eight, all right? So that means that if we extend it three bits we've given ourselves the ability to utilize or to come up with eight separate subnet IDs. Now I should say at this point that originally the formula was two to the n minus two and it was because in the classful system you couldn't have, or with classful routers you couldn't have subnet IDs that were all zeros or all ones. Well, when we extend that three bits two of the possible values would be 0000 and 111, right? So we always went two to the n minus two. With the classless system and modern routers they support zero and one subnet IDs. So the new formula is just two to the power of n. So you're simply adding bits to the subnet mask plugging it into this formula until you reach the appropriate number of subnets although you should probably always give yourself some room for growth. So once you've reached the number of required subnets we stop and then we convert the subnet mask back into its decimal form, okay? Along with that we need to make sure that the number of host bits left is adequate based on the requirements, based on the number of hosts per subnet that we had identified earlier. And then the final step would be to identify the individual subnets. Again, where the rubber meets the road as I said before, we have to actually come up with the subnet IDs. And not only do we have to come up with the subnet IDs but we have to come up with the range of addresses within each of those subnets. So this is a subnet creation process. And if this is the first time you're hearing it I can expect that you are completely confused at this point. Don't worry, that's understandable. The vast majority if not all of us were confused the first time we heard this. It is going to take looking at this process from start to finish for us to get an idea as to what we're talking about. All right, so now we've talked about binary addressing. We've looked at the process of subnetting. It's time to go back through some other fundamental concepts. And one of those would be the IPv4 address types. There are three basic address types although only two of them will we use predominantly. And so those are public addresses and private addresses. If you remember the original goal of the IPv4 protocol was to assign public addresses to all hosts that would connect to the network. That was the goal. They figured when they created TCP-IP with 32 bit IP addresses that it would provide two to the 32nd power, excuse me, number of available addresses. Well, that's 4.3 billion addresses and the developers of TCP-IP thought that that would be enough. It was not, not even close. And we've run out of IPv4 addresses. Hence our upcoming discussion on IPv6. But public addresses are globally unique addresses. These are addresses that are assigned by the IANA to registrars, like an ISP, which then assigns them to companies. Public addresses would be routable on the internet and are required for direct communication on the internet. So if you have a web server, for instance, that sits out on a perimeter network, it's going to have a public address. So that clients on the internet can connect directly to that address. On the other hand, the other type of address that you have is a private address. Private addresses are not routable on the internet. They cannot be connected directly. They would require network address translation, or NAT, which is a particular protocol that can be implemented and is implemented by the vast majority of organizations. These are assigned directly by the organization to clients. They don't have to be registered through the IANA. And technically, you can have, you can deal with multiple organizations that use the exact same address ranges. In fact, you most often will. There are three private address ranges that exist for IPv4. They are 10-000-8-172-16-00-12 and 192-168-00-16. Okay, now when we see those CIDR notations, I think the first thought is, wait a minute, those aren't the default subnet mask. You would be correct. 172 is a class B. And so the default subnet mask would actually be a 16-bit mask. Well, it's not the whole 172 range, or a whole 172-16 range that's reserved. The reserve range is technically 172.16-31. Okay, so it's 12 bits. On the class C range, it's just 192-168-00. That is the class C. The vast majority of organizations will utilize these private addresses. They're just more flexible, they're easier to use, and it comes down to the fact that, well, even though it was the goal of IPv4 to connect everybody directly to the internet, it's just not required. I don't need to connect individual client computers to the internet. I mean, I do need to allow them to have a connection to the internet, but they don't have to be connected directly. A network address translation device, in short, is simply a device that has two interfaces. One has a public address, the other has a private address. It's used as the gateway for all devices on the LAN. Those devices will, when they need to communicate with the internet, they'll send packets to the network address translation device, which can be a server, a router, a firewall, et cetera. That device will then use its own public address to go out to the internet and retrieve internet web pages, for instance, on behalf of the client, so it performs translation. It's not using the private address as the source address, it's using its own public address as the source. Now, the third type of address is APIPA, or Automatic Private IP Addressing. 169.254.x.x, and the xx means it doesn't matter what's in the last two octets. This is not an address range that is typically assigned to clients. That's why I said you'll use two out of these three. It is not assigned because it's not routable. It does not define a default gateway. It's a private range that was reserved by Microsoft for use with their operating systems. Essentially, APIPA is there as a troubleshooting mechanism. So if you have a client that is designed to obtain an IP address automatically via DHCP, and for whatever reason the DHCP server is unavailable, then that client will generate a 169.254 address. It is quite possible for you to configure multiple clients to obtain IP addresses dynamically and not install DHCP on your network. All of those clients would register or would try to communicate with DHCP, would not be able to, and then would generate a 169.254 address. When they generate the address, they would ensure that that address was not already in use on the network, and they would all be able to communicate locally. Now nobody does that, trust me. Nobody does that just because it's sort of senseless in that particular scenario. What this ends up being is a troubleshooting mechanism. So we typically have clients that are supposed to obtain IP addresses from DHCP. So you've got 50 clients on your network and 49 of them have valid DHCP obtained addresses, and one of them has a 169.254 address. And that one is unable to communicate with anybody else on the network. And so when we go to a machine and we see a 169.254 address, it immediately tells us that there's a problem. This machine was unable to obtain an address from DHCP. And there can be various reasons why that might have happened. The point is that it did happen, all right? And so it's an automatic giveaway. And I'm telling you that because in real life, if you see this or the exam you see this, if you ever see a 169.254 address, it's an automatic indication that this client did not obtain a valid address from DHCP. So we know exactly where to look to start troubleshooting the situation. All right, so there are three address types, but in reality, public and private addresses are the choices. Public addresses are used for machines that need to be directly accessible from the internet, of which most organizations are only gonna have one or maybe a couple, and private address ranges are used for all other systems. There are also a couple of different types of IPv4 networks that we're going to encounter, simple and complex. Now these are just general categories, but a simple network will use classful private IP addressing for internal host, okay? We use the term classful because we mean that subnet mask are gonna be the same for all subnets and the subnet mask are gonna stop on the decimal points. When subnet mask stop on the decimal points, it makes life a whole lot easier for us because we don't have to do binary math, essentially. We know that if an address is 192.168.1.200 and it's got a 255.255.255.0 subnet mask, then we know the first three octets are the network ID and the last octet is the host ID. I don't have to convert into binary or do any crazy stuff. I just know that's the case because the subnet mask is stopping on the decimal points. It's also much simpler when all of the subnets have the same length of subnet mask. When you start getting into more complex network scenarios, you'll be using class less, excuse me, addressing. And you could be using private or public, although the private addressing is still far more common. The key here, though, is that subnet mask are created based on the required number of subnets and host. So we've extended the subnet mask, we've performed this process of subnetting, but we don't necessarily have to stop the mask on the decimal points and we don't have to make every subnet mask the same length. You're going to have certain subnets that only need two addresses, like a subnet that connects two routers, for instance. And so this is referred to as VLSM or Variable Length Subnet Masking. And it's advantageous when there are huge differences that exist between subnets in relation to the number of hosts. You can almost guarantee you're going to see these kinds of scenarios, both in real life, as well as on the exam. So we need to understand exactly how this process works. MUSIC