What is a network?
A network is really just a really fast way of moving
information around. When you are working
as a network engineer, your job is of course to troubleshoot and monitor configurations
& implement changes; but beyond this, your job is to keep communication on
your network as fast and efficient as possible!
A network is nothing more than the infrastructure to
transfer information (resources).
Who is Cisco?
Cisco is one of the world leaders in network equipment. Their claim to fame, is Cisco IOS. This stands for “Internetworking Operating
System”. This is the modular operating
system (software) that runs on top of most of Cisco’s networking equipment and
has become the standard for routing and switching. Other vendors have gone as far as
implementing many of the features that are in IOS.
Networking Diagrams,
Devices and their purposes.
For each of the devices that I learned about, I am going to
show two images of the device. One is the logical representation of the device, and the other is an
example image of an actual Cisco device.
Typically the logical representation is what you'll see in a network diagram. It’s more cartoon like, but the beauty of this is that it is
not specific to the model or even brand of device that you’re using. You could take almost exactly the same
logical networking diagram, and deploy it exactly as depicted using Cisco, Brocade or Juniper equipment.
Switches:
Switches provide the interconnecting points for your local area network (LAN). A switch typically has a very high port density (lots of ports), as can be seen in the picture above. A switch is not useful for communicating outside of your local area network. Their primary job on your LAN is keeping a MAC address table.
Routers:
Routers provide the ability to communicate to the wide area network (WAN) the ability for joining networks together, or for communicating with systems that are not on your local area network. The “internet”, is really nothing more than (a lot of) routers connected together. It’s essentially a network of networks. Routers mark the dividing line between your network and another network. Important to note, that one of the primary functions of a router is to terminate broadcasts. This is important to know, and I’m sure we’ll get into this in a later post, but understanding this can help with troubleshooting at some point, so I thought I’d throw it out there. Another very important task that a router must do is maintain a routing table, so it can know how to get you from point A to point B.
Wireless Access
Points:
I have seen a variety of icons for these, and quite
honestly, I’m not sure that we need to look deeply into these for the CCNAX
exam. With that said, it’s important to
note that in a lot of ways, using a Cisco or any other brand of wireless access
point at this time is far less efficient than using a switch. A switch limits the collision domains (more
on this later), to a single domain per port, whereas a wireless access point
does not, leading to inefficiencies and it acts much more like a hub (more on this later).
Endpoints
Here you’ll see, PC’s, Servers, printers, etc. You can also see things like firewalls, hubs,
different types of connections, phones, layer 3 switches and other devices.
Network Diagram Example vs. Real World
 |
| Logical Network Diagram |
 |
| Physical Network, in real life. |
Connections &
Speed: How fast is that?
Speed in computers is measured in bits, not bytes! This is an important distinction, primarily
because when your operating system is downloading a file, it’s typically going
to show the download in MB (megabytes), which is a measurement in BYTEs, not
BITs. A single byte, equals 8 bits. Storage is typically measured in bytes, which is where
so much of the confusion comes in.
A bit is just an on (0) or an off (1). This is the ones and zeros that are often
referred to when you’re thinking of information on the computer.
So, when you are paying for a 10 Mb (megabit) connection,
and you go to download something that is 100 MB, you will very readily see the
difference as instead of taking what some people expect to be about 10 seconds,
it actually ends up taking about 80 seconds.
This is a source of confusion if you don’t know what you’re looking
at! Again, this is all about being able
to walk the walk and understand the discussions that are going to start right
away when you’re starting to talk to other networking people.
To figure out the amount of time it will take to download a
file, the easiest thing to do is to convert the file into megabits, and then do
the math, so that you’re then comparing like values.
Types of Connections
Ethernet – Most common in a LAN. CAT5. Common
speeds today include:
- 10 Mbps
- 100 Mbps
- 1 Gbps (aka 1000 Mbps)
- 10 Gbps (aka 10000 Mbps)
- 40 Gbps (aka 40000 Mbps)
- and on and on...
Fiber – Super fast, high speed connections, can be very long. Also, these are common in a LAN for very high
bandwidth connections. This is
Serial Cables – Still used to bring up the internet
connection.
Talking the Talk: The OSI Model & the TCP Model
This is one of the most important areas to really understand as someone who is new to networking and Cisco. The OSI model is not only where you really learn to talk the talk, but when things start going wrong on your network, and your helpdesk shouts across the room in a panic that "The internet is down!", this is going to be step one in trying to not just diagnose what the real problem is, but also to try to translate "the internet is down" into some useful information. One diagram that I really like is below. The copyright and attribution is contained within the image and I'm leaving it because so much information is included in this one diagram. Refer back to this often as I go into the OSI Model and the TCP Model. There are also some other diagrams that I'll post at the end of this post which show alternative depictions of this model for your review. Find one that works for you and learn it, inside and out. You'll be a better network engineer for it.
This is one of the most important areas to really understand as someone who is new to networking and Cisco. The OSI model is not only where you really learn to talk the talk, but when things start going wrong on your network, and your helpdesk shouts across the room in a panic that "The internet is down!", this is going to be step one in trying to not just diagnose what the real problem is, but also to try to translate "the internet is down" into some useful information. One diagram that I really like is below. The copyright and attribution is contained within the image and I'm leaving it because so much information is included in this one diagram. Refer back to this often as I go into the OSI Model and the TCP Model. There are also some other diagrams that I'll post at the end of this post which show alternative depictions of this model for your review. Find one that works for you and learn it, inside and out. You'll be a better network engineer for it.
 |
| Come back to this often. If the copyright holder accepts donations, give them. |
It is important to know not only the layers, their names and order, but also what they do, what devices communicate on which layers and how they do so, which is why I absolutely love the different diagrams that I've shown in this post! One mnemonic device that I am going to use to remember the layers of the OSI model is the following:
Please Do Not Throw Sausage Pizza Away
The first letter, of course referring to the first letter of each layer 1 to 7 in the OSI model.
What's in a number (and a networking joke that you'll probably hear)?
Just to call attention to the layer numbers, these are typically how network engineers describe problems that they are seeing on a network and is something that you will likely hear often.
- "It's a layer 3 problem, maybe the routing table is wrong?"
- "I suspect that it's a layer 1 problem, could you have someone go ahead and verify the cable is physically plugged in?"
- "We've been working this all week. Clearly this is a layer 8 problem."
How the communication happens in the OSI model.
Before going any further, a general description of how the transfer of data between two systems happens and what each layer of the OSI model is doing to the data as it goes through the OSI model. A diagram that does a pretty good job of showing that is listed below:
Things may get a bit tricky as I describe the flow of data through this to you, so if you have any questions or confusion, refer back to the diagram above to understand logically what's going on. For ease of understanding, note in the diagram that each step of the way along down the osi model, the data is having bits of information added to it, and each step up the osi model on the other side, those same bits of data are being removed. This is called encapsulation. Encapsulation is the process of adding bits of data onto the data in order to guide the data to where it is intended to go.
First, we'll go through the layers, what they do and hopefully keep things clear, then I'll try to provide an example to demonstrate the process of encapsulation.
The Layers
7 - Application Layer - This is the layer or the API's, that is built into the Operating system that the application interfaces with if the application needs to send data over the network. This isn't the application itself, but more often the protocols that the application would use, such as ftp, http, www, telnet, ssh, etc.
6 - Presentation Layer - This formats the data into a well known format to prepare it for transport. Basically treating the data as some form of generic data. Examples of this would be .jpg or .gif for images, or .html for a web request or mp3 and wav for audio. Encryption is also normally done at this layer if it is encrypted by an application and not a physical device on the network.. Typically this is going to include the extension or type of file.
5 - Session Layer - This manages the connection between the local and the remote application. This starts and ends sessions, creates session id's etc.
4 - Transport Layer - Ensures the integrity of the data transferred. This is where TCP (reliable) & UDP (unreliable) is determined.
TCP - Transmission Control Protocol
UDP - Unreliable Data Protocol
This is also where ports are assigned, both source (a random, high numbered port assigned by the system) and destination (normally a predetermined port based upon the type of communication). The destination port, is also quite commonly referred to as a common port and is between 1-1024. You certainly do not need to memorize all of them, but there are certainly some that you will want to remember. A nice cheat sheet for the common ports can be found here.
The source port, which is normally between 1025 and 65535. This is assigned dynamically by the operating system. You can get an idea about the source ports in use by opening a command prompt on a windows system, or even a mac and running the command 'netstat'. The results of netstat can be seen below:
The source port, or in the screenshot above 'local address', is how your computer determines where the information that is returned to your system knows where to go to be properly received.
3 - Network Layer - Logical addressing is addressed here. This controls how the data gets from one host to another. Here, we are referring to IP Addresses. This is the land of packets, subnetting, path determination & addressing. This is also the land where routers live.
2 - Data Link Layer - Physical addressing is handled here. Here, we are referring to MAC addresses. This is the land of frames. This controls how to get data from one location to another location on the LAN. This is where switches & bridges live.
1 - Physical Layer - This is the actual physical connection. This is ones and zeroes, electricity and such. This is the actual cabling, hubs, patch panels, etc.
The Example
So in this example, we'll say a cat named "Sonny Boy" (this is the name of my daughter's cat, and he looks a lot like the cat in this example) is using a web browser and wants to watch some video of some mice on youtube! As he is using a web browser, the data that he is going to be sending is going to be an http request in this case. A diagram to show the encapsulation of the data is shown below:
As the data moves down the stack, that's when encapsulation is happening. At each layer, a little bit of information is tacked onto the data, which is what allows the data to reach its final destination, well sort of. Let's talk it through...
At layer 4, the first 'networking' layer, the data is given the TCP Seg Payload. As this is an http request, the payload is likely TCP. (If this were a request for the actual streaming video, it is much might be UDP.) Then, before the packet takes off to the next layer, there is also a source port (high) and a destination port (normally common) added to the data. The ports are used to identify where the traffic needs to be sent when it gets to the host, and when it gets sent back to the system.
For my own visualization, I tend to think of the data being sent to the PC, like mail being sent to an apartment building. The tag is like the apartment number. At this point, the entire encapsulated data is called a segment.
At this layer, the NIC, which is a layer 4 device, will utilize DNS to determine the IP address of where this needs to get to. DNS, which I will likely cover in more detail later, is the Domain Name Service, and it is what translates what you and I see, such as blog.trivialpackets.com, into an IP address. So at this point, a source ip address and destination ip address is added to the segment. These denote both the starting and final ending point for your data. At this point, the encapsulated data is called a packet.
But we're not done yet. One last thing needs to be assigned before we release this out into the wild of our network. The system will know, based upon subnetting, ip addressing and other information available, whether or not the destination IP address resides on the same network or not. I'm sure I will go into this later, in depth, as it is important. Based upon this knowledge, the data will be encapsulated further to include both a source and a destination MAC address. This is a hexadecimal address tied to a network component. The source MAC address is the MAC address of your computer's NIC, and the destination MAC address will be one of two addresses, depending on whether or not the destination IP resides on your network:
Full disclosure: I am only unsure as to whether this all occurs on the NIC card, or if the actual encapsulation at the network and data link layers. If it does, I'm not 100% certain how it would know the information if it happens after the NIC. Hope to learn this as I go on. If you know the answer, feel free to comment, otherwise I'll update when I know for sure. Either way, I won't hold up posting this any longer.
A diagram of what the specific payloads are, and when they are added to the data that is being sent can be seen below: Notice that after layer 4, the data plus the information is referred to as a segment. After layer 3, the data begins adds the source and destination IP ports and the entirety becomes known as a "packet."
The data is then sent forward to the switch. The switch's job is to basically remember which mac addresses are located on which ports, so when the frame gets to the switch, it sends it out to the port of the destination mac address. If the data is destined for something internal, it would go to that port, but in this case, we're going to YouTube, so it would go to the port that leads to the router.
This is where it kind of gets interesting. The router will tear open the frame, because it was addressed to the router. The router will then examine the packet, determine if the packet belongs either to the router (has the ip address of the router) or belongs to an ip address on the network. If it does not belong to the same network, the router will use the internal routing table to send this packet off into the internet as a frame again, however this time, it will use the mac address of the outside interface as the source mac address, and will use the destination mac address as the next router in line based upon the routing table. This tearing off and replacing of the mac address information in the ethernet frame continues until a router discovers that the IP address does belong to the same network and will pass the packet on to the switch on the destination network for delivery.
The destination switch will look at the IP address and pass the information off to the host. The host then takes over, and directs the data, using the port source and destination ports, the transportation type and get it to the appropriate application on that host.
Keeping in mind that this happens every single time any tiny, tiny piece of data flows across the network can make you really appreciate exactly how fast our networks really are, even when the page is "running slowly".
First, we'll go through the layers, what they do and hopefully keep things clear, then I'll try to provide an example to demonstrate the process of encapsulation.
The Layers
7 - Application Layer - This is the layer or the API's, that is built into the Operating system that the application interfaces with if the application needs to send data over the network. This isn't the application itself, but more often the protocols that the application would use, such as ftp, http, www, telnet, ssh, etc.
6 - Presentation Layer - This formats the data into a well known format to prepare it for transport. Basically treating the data as some form of generic data. Examples of this would be .jpg or .gif for images, or .html for a web request or mp3 and wav for audio. Encryption is also normally done at this layer if it is encrypted by an application and not a physical device on the network.. Typically this is going to include the extension or type of file.
5 - Session Layer - This manages the connection between the local and the remote application. This starts and ends sessions, creates session id's etc.
STOP!
Note that everything above this point happens on the computer or device itself. Everything below this point is happening on the network and what we really need to dig in and focus on as aspiring network engineers!
As you can see in my favorite diagram above, everything that occurs before the network is the "DATA" on the encapsulation diagrams. Some diagrams, like the one just above, show encapsulation occurring at every step along the way. This is accurate in the OSI model, and is important to note. However things really start to get interesting from a networking perspective from here on out. If something happens above this point, it's probably someone else's issue, so just bear that in mind.
4 - Transport Layer - Ensures the integrity of the data transferred. This is where TCP (reliable) & UDP (unreliable) is determined.
TCP - Transmission Control Protocol
UDP - Unreliable Data Protocol
This is also where ports are assigned, both source (a random, high numbered port assigned by the system) and destination (normally a predetermined port based upon the type of communication). The destination port, is also quite commonly referred to as a common port and is between 1-1024. You certainly do not need to memorize all of them, but there are certainly some that you will want to remember. A nice cheat sheet for the common ports can be found here.
The source port, which is normally between 1025 and 65535. This is assigned dynamically by the operating system. You can get an idea about the source ports in use by opening a command prompt on a windows system, or even a mac and running the command 'netstat'. The results of netstat can be seen below:
The source port, or in the screenshot above 'local address', is how your computer determines where the information that is returned to your system knows where to go to be properly received.
3 - Network Layer - Logical addressing is addressed here. This controls how the data gets from one host to another. Here, we are referring to IP Addresses. This is the land of packets, subnetting, path determination & addressing. This is also the land where routers live.
2 - Data Link Layer - Physical addressing is handled here. Here, we are referring to MAC addresses. This is the land of frames. This controls how to get data from one location to another location on the LAN. This is where switches & bridges live.
1 - Physical Layer - This is the actual physical connection. This is ones and zeroes, electricity and such. This is the actual cabling, hubs, patch panels, etc.
The Example
So in this example, we'll say a cat named "Sonny Boy" (this is the name of my daughter's cat, and he looks a lot like the cat in this example) is using a web browser and wants to watch some video of some mice on youtube! As he is using a web browser, the data that he is going to be sending is going to be an http request in this case. A diagram to show the encapsulation of the data is shown below:
As the data moves down the stack, that's when encapsulation is happening. At each layer, a little bit of information is tacked onto the data, which is what allows the data to reach its final destination, well sort of. Let's talk it through...
At layer 4, the first 'networking' layer, the data is given the TCP Seg Payload. As this is an http request, the payload is likely TCP. (If this were a request for the actual streaming video, it is much might be UDP.) Then, before the packet takes off to the next layer, there is also a source port (high) and a destination port (normally common) added to the data. The ports are used to identify where the traffic needs to be sent when it gets to the host, and when it gets sent back to the system.
For my own visualization, I tend to think of the data being sent to the PC, like mail being sent to an apartment building. The tag is like the apartment number. At this point, the entire encapsulated data is called a segment.
At this layer, the NIC, which is a layer 4 device, will utilize DNS to determine the IP address of where this needs to get to. DNS, which I will likely cover in more detail later, is the Domain Name Service, and it is what translates what you and I see, such as blog.trivialpackets.com, into an IP address. So at this point, a source ip address and destination ip address is added to the segment. These denote both the starting and final ending point for your data. At this point, the encapsulated data is called a packet.
But we're not done yet. One last thing needs to be assigned before we release this out into the wild of our network. The system will know, based upon subnetting, ip addressing and other information available, whether or not the destination IP address resides on the same network or not. I'm sure I will go into this later, in depth, as it is important. Based upon this knowledge, the data will be encapsulated further to include both a source and a destination MAC address. This is a hexadecimal address tied to a network component. The source MAC address is the MAC address of your computer's NIC, and the destination MAC address will be one of two addresses, depending on whether or not the destination IP resides on your network:
- If the destination IP address resides on your network, then the system will perform an arp broadcast to find the MAC address of that host and use that as the destination MAC address.
- If the destination IP address does not reside on your network, then the system will perform an arp broadcast to find the MAC address of the router (AKA default gateway) and use that as the destination MAC address, knowing that the default gateway will know what to do! The router will know what to do, as one of it's primary responsibilities is to contain a routing table, so it knows how to get where you need to go.
Full disclosure: I am only unsure as to whether this all occurs on the NIC card, or if the actual encapsulation at the network and data link layers. If it does, I'm not 100% certain how it would know the information if it happens after the NIC. Hope to learn this as I go on. If you know the answer, feel free to comment, otherwise I'll update when I know for sure. Either way, I won't hold up posting this any longer.
A diagram of what the specific payloads are, and when they are added to the data that is being sent can be seen below: Notice that after layer 4, the data plus the information is referred to as a segment. After layer 3, the data begins adds the source and destination IP ports and the entirety becomes known as a "packet."
The data is then sent forward to the switch. The switch's job is to basically remember which mac addresses are located on which ports, so when the frame gets to the switch, it sends it out to the port of the destination mac address. If the data is destined for something internal, it would go to that port, but in this case, we're going to YouTube, so it would go to the port that leads to the router.
This is where it kind of gets interesting. The router will tear open the frame, because it was addressed to the router. The router will then examine the packet, determine if the packet belongs either to the router (has the ip address of the router) or belongs to an ip address on the network. If it does not belong to the same network, the router will use the internal routing table to send this packet off into the internet as a frame again, however this time, it will use the mac address of the outside interface as the source mac address, and will use the destination mac address as the next router in line based upon the routing table. This tearing off and replacing of the mac address information in the ethernet frame continues until a router discovers that the IP address does belong to the same network and will pass the packet on to the switch on the destination network for delivery.
The destination switch will look at the IP address and pass the information off to the host. The host then takes over, and directs the data, using the port source and destination ports, the transportation type and get it to the appropriate application on that host.
Keeping in mind that this happens every single time any tiny, tiny piece of data flows across the network can make you really appreciate exactly how fast our networks really are, even when the page is "running slowly".
More pretty pictures please!
As promised, some alternative OSI Model diagrams. I will add more as I find anything useful!
 |
| Link. Also a pretty cool article... :) |
 |
| Couldn't find the link, sorry! |
