Networking Basics

22. Transport Layer Concepts Part 3

It handles End To End connections between two systems. It also handles transportation issues between two hosts. It also ensures data transport reliability and establishes, maintains, and terminates virtual circuits. It provides reliability through fault detection and recovery via retransmission, and it employs the segment as its transmission unit technologies that we discussed earlier, as well as the distinction between TCP and UDP.

We also talk about the TCP three-way handshake where the client sends an initial TCP scene and as the server receives it and the ports are open, it will reply with a TCP flag The client will finalize the handshake by sending an acknowledgement flag. These are the flags that we talk about. We also talk about some of the most common flags, such as not only scene adapt but also pushing reset unfit flags. We also talk about windows, either sliding windows or window sizing. This allows us to know if the server can handle the data that is sent by the clients, because our goal is to ensure the sender is not overwhelming the receiver by sending more data than it can handle. So this is the concept of flow control.

Take note that the larger the window size, the better, as it indicates that the network is of higher quality and the server has more available resources. We also talk about retransmission. When the server didn't receive all of these messages, it will send an acknowledged number and the client will assume that some of these messages were not received by the server. So the client will do the retransmission. TCP versus UDP transport TCP is a connection-oriented protocol, while UDP is a connection less. TCP does error checking and UDP does it. TCP is reliable, while UDP is an unreliable protocol and doesn't guarantee data delivery. TCP does ordering and sequencing, while UDP sends packets in any order. TCP is slower on UDP because TCPestablish connections, thus error checking and many more. These are the applications that run unused by TCP and UDP. Under TCP, we have our Web based application Http andHttps running on ports 80 and four three. We also have remote access applications such as Telnet and SSH ports 23 and 22. We also have FTP email applications such as SMTP and Ixopo three.

Under UDP we have DNS boot p. With DHCP, we also have TFTP, SNMP, and many more. We've already completed discussing the lower layers of our OSI model; the physical layer, the data leak layer, the network layer, and the transport layer. Now, if we compare this to our TCP IP layer, it's almost the same except for layer one and layer two. These two layers are consolidated in the TCP/IP model into only one layer, called the network interface layer. And it makes sense because when we talk about the network interface guard, it's both layer one and layer two. We also have the network layer for our OSI and for TCP IP.It's called the Internet layer, which is basically the same when it comes to functionality. It's just that the names are different. But the transport layer in the OSI model and the transport layer in the TCPI model are 100% the same, even the name. Now let's review all of the lower layers that we've discussed, because in the exam there will be a higher chance that this will be asked. First, the layer. One physical layer. It's all about electrical and mechanical things that are running on our network devices and hosts. And if you see transmissions, if it's bits or binary, it's all about the physical layer. If you see duplex and speed, these are Port configuration.

And again, it's all under physical layer cables modules such as FFP plus Copper and Optical. This is part of the electrical and mechanical hardware and physical All of this stuff is running on the physical layer. Next up is our data link layer, or layer two. In Layer Two, this is where we encapsulate and de capsulate network data pockets into frames or vice versa. We also use this for transforming a raw transmission facility into an aligned undetected transmission facility to the network layer. And if you see these terms and configurations such as VLAN 802, One Two, LACP, or Trucking ARP Mac Address, especially the frames as a transmission unit, this is all running in the Data Link Layer. Layer Two Next up is the network layer, and this is where we forward packets and route them to intermediate Layer Three devices such as Routers okay.

The network layer also handles layer three. Flow controls, error controls, packet sequence fragmentation, ICMP TTLs. All of these operate in layer three, or the network layer. When it comes to addressing, the data link layer uses a Smack address, our layer three or the network layer uses an IP address, and there are many types of IP address. We already talked about this. We have the local host, we have Unicast, we have the Anycast broadcast as well. Layer Two also has a broadcast. However, in Layer Three, we use IP Addresses, and in the Network Layer, we also use routing. And we have two options: static routing protocols and dynamic routing protocols such as OSPF and BGP. Next is the transport layer. We just talked about the transport layer, but I'm just going to repeat it and highlight the terms that you need to take note of. We have connection oriented and reliable flow control windowing, window size retransmission, and again, it's responsible for End-to-end communication over a network segmenting data for manageability and reassembling segments. Data and TCP is all about the TCP three-way handshake, sequencing, and acknowledgement between two hosts.

23. Network Address Translation (NAT)

Now you access the network whether you're using wireless or wired device, there's an IP address assigned and you can reach the network. You can reach devices such as servers, laptops, default gateways, and even printers. How about if you can access the Internet? Well if you do, are you using the same IP address? Probably not because probably you're using an RFC 1918 or the private IP address which is not allowed to route to the public network such as the Internet. Once the packet is set and rerouted to your router and then to the service provider's router, once it receives it and realize hey, this source is a private IPad dress, then that service provider router will drop it. Now, the solution to that is a translation. We call it Network Adverse Translation? Or Simply Not. Not allows private users to access the public network, which is the Internet, by sharing one or more public IP addresses. And our goal is to use our devices to access the Internet. Nat has three types. The first one here is one to one address mapping. Now when I say one to one meaning this the first one is your private IP address and this one is the public IP address. Now we also have many to many when Isai many public IP address, this may be added or retrieved from a public IP address pool. Now, when you say many private IP addresses, this is obvious.

Our land, our local area network, may have not only hundreds of IP addresses but thousands of them. And lastly, many-to-one address mapping. This is the most common implementation of Nat. It exists everywhere, including on your home network. Now this is an example of how Nat operates. As you can see, we have our own LAN. We have three PCs with an IP address of 192 168.Numbers one, two, and three We also have a router and the router has two interface. Obviously, the other interface is Ethernet, which is part of the 191-6810 network. And the other interface is the one with an IP address of 108 171-6151. Now, here's what it's going to do. PC one must send traffic to the server using an IP address of one to two, 8199, dot two to 458. It will first send traffic or send packets to the router because that's its default gateway.

Once the router receives it, the router will start translating the traffic. So let's do it. 192-1681, same packet, the router. And look at that, the router translated to one 8171, dot 61, dot 51. What is this IP address anyway? This is the IP address of the router's IP interface. Ethernet zero, to be specific. And as you can see, if you look at the nuts table, you see the inside look at this is the PC one IPad dress, one I 2168, one with a local port of 1516. Now the inside is global. This is the translated IP address triggered by the router. And the IP address here is what? It's the same IP address of Ethernet zero. But look at the port. It's using the inside local IP address port. Which makes sense because we are just translating the IP address anyway. This is the IP address that's been used to reach this server here. So the server, if you ask him, has 192-1681 contacted you? The server will say "no, I don't know him." What the server will see is the IP address of 181-71-6151 with a port of 1516. Now let's try another example.

Let's see how PC number two operates. OK? It sends traffic to the router and the router does the same. It converted 181 68 1 2 to the same IP address 108, 171-6151. If we look at the table, everything is the same. What differs is the port 1517 on both the inside and the router's IP address. Maybe you're thinking, what if there's another network, another PC that's using the same port 1516? Can the router use 1516? The answer to that question is that no router will look for an available port because 1516 is already no router Many to many networks are now another type of network is many to many. In that case, we may not use the router interface but rather a pool configured for the router. So maybe a real router has a pool of IP addresses. And this can be a range of 12819 to nine, excuse me, this one, maybe it has a range of 181, 71, 61, let's say 55 to 65, for example. right? So this is an example of "too many" rather than "address mapping." Now, we also have one to one. How does it work? Well, if we zoom this server here and let's say it's sitting in the corporate data center, you have a router and you have a server.

This server may have an IP address, which is a private IP address. What I need to do is 168 ten, for example. This IP address is a private IPad dress, meaning it's not routable to the Internet. What we can do is translate this IP address to 128-192-2458. Why is that? so that the server becomes routable. And maybe you're thinking, why would not just directly assign these public IP addresses to the server instead of enabling that? Well, you may do that, yes, no problem. But most of the data centers today do not have just one server. Maybe this is a web server talking to a database server talking to a caching server. And these internal servers are all using private IP addresses. So the most common implementation is that all of your servers have private IP addresses and we just enable one not. This is referred to as a static map by some vendors and platforms. But for F five big IP we just call this Nap. Simply Nap.

24. Dynamic Host Configuration Protocol (DHCP)

How do you configure the IP address of your PC? Well, it's pretty simple. If you're using Windows, you just go to Control Panel Network and Internet Network Connections and choose the network adapter. Right-click Properties and choose PCP IP Four. And from there, you can change your IP address. Now, this is okay if you're in a large or small organization where you only have ten PCs or a maximum of maybe 20 PCs. But if you're in a large environment and you're not just dealing with PC Desktops or you're also dealing with a lot of wireless devices, maybe static or manual IP Address Assignment is not the best for you.

Why is that? Well, because we have a lot of disadvantages, as you know, with static or manual configuration. You need to do this per network adapter. So, if you have three network adapters in your PC and want to connect them to the network, you must configure each one separately. It's time-consuming and prone to errors. This is very obvious and extremely rare for a wireless network environment. Have you ever seen someone configure IP addresses on their smartphones or even on their tablets? No. You won't see that it's extremely rare, maybe for testing, in a normal environment, a normal case, no way. Now, there's a better option. This is what we call the DHCP IP address assignment. It is good because first it is centralized and IP addresses are allocated and distributed, and then it pushes them automatically to the clients. And it's always per VLAN settings or per Broadcast Domain because DHCP traffic the request reply.

Everything is a layer of traffic. all right? But there are also ways for the ACP replies to be forwarded between client servers in different subnets or different broadcast domains. This is what we call a DHCP Real Estate Agent. And many network devices can enable this, okay? Because DHCP doesn't see two oldVLANs, maybe just one or two. But if I have ten or twenty VLANs, I just need to enable or activate one or a couple of DHCP servers and all of the VLAN's IP addresses will be allocated to that DHCP server or from that DHCP server. So how does this dynamic host configuration protocol, or DHCP operating? It’s very simple. What we have here is the PC and it's connected to a network together with a DHCP server. Assuming it does not yet have an IP address and the network adapter is configured to the ACP client, also known as Automatic IP Address Assignment, it will begin the first step. What is the first step? It will send Discover or DHCP Discover Message, which is a broadcast traffic. Once the DHCP server receives it, it will either respond or reply via unicast. We call this an offer message. Now, an offer message is just simply telling the PC client, "Hey, Mr. PC, I have this IP address 170 216-1." Do you want to take it?

Now, if the PC wants to get that IP address, it will simply send the Http request, which is another broadcast traffic. Once the server receives it, it will acknowledge it and provide the IP address to the client. Now, if you look at the DHCP message type, we have more than four. We have four steps, 1234. Just say Note that the message type has different numbers. For example, all right, the discover is message type one. The DHCP offer is message type two. The DHCP requests message type three. But step four doesn't match the DHCP message type.

DHCP message type four is DHCP decline. But in our example, we use DHCP message type five, which is DHCP. Acknowledged Now, assuming everything worked properly, the result would look like this. Your PC now has an IP address of 170216 with a net mask of 245-245-2550. Plus, the router default gateway list time is one day on the DNS server. 170, 216, 1254. Why does DHCP not only provide an IP address but other parameters as well? This is what we call the DHCP options subnet But when it comes to configuration, it is mandatory when it comes to but routers or DNS server. This is not mandatory, but in a real-world scenario, this is very common. The HCP server is not configured by IP address and net mask. Most of the time, they also add the full gateway or the router plus the DNS server. Or it can have multiple DNS servers. In addition, the IP address list time.

Prepared by Top Experts, the top IT Trainers ensure that when it comes to your IT exam prep and you can count on ExamSnap Application Delivery Fundamentals certification video training course that goes in line with the corresponding F5 101 exam dumps, study guide, and practice test questions & answers.