We’ve previously talked about the OSI model but did you know that it’s not the only model out there its competition is the TCP/IP model in this article we’re going to learn about this model and see its perspective on the network if you.
Haven’t reading the OSI article yet now is the time to go back and check it out as we discussed there a model is a set of layers that describe how networks hardware and software work together the goal behind the OSI model was to allow different protocols and vendors to work together the idea was to build protocols directly based on this model, but we don‘t see a lot of them today the main reason is the protocols from the competition. the TCP/IP model gained more popularity the TCP/IP framework was originally developed by the US Department of Defense later various universities and other parties got involved this led to rapid development which is probably why it has been more successful than OSI in a practical sense now we have protocols to like TCP UDP and IP which align directly to this framework another reason this framework has been so successful is that it’s willing to use existing protocols like Ethernet if they already exist it doesn‘t try to invent the wheel any protocol that is created as a part of the framework is defined in an RFC document maybe you haven’t come across these yet RFC means request for comments and is a very technical explanation of how the protocol or concept works some of these have been around for decades and are still used today as these documents are available to everyone any vendor can create hardware and software using these specifications they’ll all work together right now we’re looking at RFC 1122 which is the original description of the TCP/IP model released way back in 1989 now let’s take a look at the TCP/IP model a little closer just like OSI this model is made up of several layers you can also see there are two TCP/IP models which I’ve listed here as original and current the original model had four layers and was defined in RFC 1122 which we looked at Earlier, the revised model splits the bottom layer into two separate layers which we’ll talk about soon the fantastic thing about the current version of the TCP/IP model is that it lines up well with OSI you can see that the OSI session presentation and application are all rolled into a single application layer personally I prefer this I like to think of this entire area as the application rather than trying to break it out like OSI us I also personally think that we can break these models into two main areas the top half is the application and their processes in the application layer we’re talking about protocols like HTTP for web browsing SMTP and IMAP for mail delivery and FTP for file transfers these applications create processes which listen on particular ports that’s the job of the transport layer which uses the TCP and UDP protocols we’ll dig deeper into those very soon to the bottom half focuses on getting the data from one host to the other the network layer uses the Internet Protocol or IP you should be pretty familiar with this from the last two, articles the physical and data link layers move the data from device to device that is across switches access points routers and other network equipment Ethernet is a very common protocol here and on this we can see that the new version has broken the original link layer into the data link and physical layers in the old days there weren‘t very many physical connection options now we have Ethernet Wi-Fi fiber-optic cables, and so many more well it originally didn‘t make a lot of sense to break up their physical connections from the data delivery now we really have to and that’s why we now have two layers here instead of one now into some more detail starting at the top the application layer this layer defines the communications between applications on two, hosts this doesn‘t define how the applications work of course instead it describes on how the applications use the network more examples here include ssh telnet and RTP for a motor administration, but we’re just going to focus on HTTP as an example because even if you don‘t know it you’re already familiar with it we start with the applications these are the web browser on the client side and the web server software on the server side HTTP is the protocol that the applications use to communicate if the browser wants to get a web page from the server it will send an HTTP request just like OSI the request passes down through the layers on one side, and back up through the layers on the other side logically one layer talk to the corresponding layer at the other side the server will go through the same process to respond with the requested web page the transport layer creates and maintains conversations between application processors on hosts this is where TCP, and UDP protocols come in we haven’t discussed these a lot yet but put simply they use port numbers to track sessions let’s continue using HTTP as our example the web server may have a process listening on port 80 when the client prepares its HTTP request it will add a TCP header we saw headers previously in the OSI article as well the TCP header contains among many other things port 80 as the destination it will also choose a port to send the request from after the header has been added we call this entire piece of data a segment when the server responds port 80 is now the source and the clients port is the destination the combination of port numbers along with IP addresses enables systems to have more than one session open at a time and to track them without this you couldn’t have multiple tabs open in your browser, for example there’s a lot more to tcp, and we haven’t even mentioned UDP yet, so we’ll have a look at them in a bit more detail in the next article now back to IP addresses we’ve covered them along with routing and submitting in details in the last article when the transport layer is done the data are passed to the network layer where it is broken into packets which are manageable chunks of data we may end up with a single packet or many packets it all depends on how many data there be to transfer the point of this layer is to make sure that data from one host can find its way to another host and to do that an IP header containing the source and destination IPS is added to every packet the path from one host to another may cross many networks packets need a router to pass from one network to another that’s why it’s called routing when the packets arrive the host strips the IP headers off, and reassembles the packets into the original data before it’s passed back to the transport layer other networks layer protocols you should be aware of our ICMP and up but IP is enough for us right now, so we’ll discuss the other two another time now,
it’s time to quickly check that you’ve been following along let’s says that you have two hosts on two different networks, and they need to communicate.
what device makes as possible and as usual?
the answers are on the website the data link layer is responsible for delivery of traffic on a single Network segment or LAN in TCP/IP terms this means delivery within a single subnet common protocols here are Ethernet and point-to-point protocol Ethernet is very common, so we’ll use that as our example an address called a MAC address is assigned to every device other protocols will use some form of dressing as well as to add these addresses we add another header in the case of Ethernet a trailer is added as well this entire piece of data is called a frame if the two hosts are in the same subnet then delivery is a simple one host ascends to another but if they are in different subnets we need a router that means that the destination MAC will be the MAC of the router itself when the frame reaches the router works out where it needs to send the packet by looking at the destination IP in the IP header it then sets its own MAC as the source and the next device as the destination if there are several routers in the path the MAC addresses in the frame will be rewritten several times along the way.
what are the names of each piece of data when they are encapsulated in TCP/IP and Ethernet?
the physical layout is exactly what the name suggests it’s responsible for physically transmitting and receiving data there are many ways this can be done Electrical radio and light signals of the options that come to mind in any case the data is encoded and the bits are transmitted over the medium this will likely happent a few times along the path like copper from workstation to a switch fiber between switches and copper through to the destination let’s check.
now how much you’ve learned how many headers and trailers will the original HTTP request being capsulated in by the time it reaches the physical layer?
what are the primary differences between the OSI and TCP/IP models?
what are common protocols that are used at the datalink network and transport layers?
if you are studying for the CCNA exam have a look at the official study, guide I’ll include a link in the description for you in the next article we’re going to have a look at TCP and UDP I hope you liked this article please let me know what you think in the comments.