The OSI Network Model - What You Need to Know

In computer networking you've got to take a layered approach to solving problems on more advanced and complex networks. It makes it much easier when you focus on one small aspect of a network at a time than focusing on the entire thing all at once.

A computer may not be able to communicate with the server and there could be any number of reasons for this. But how many physical reasons could there be? Well it could be a cable unplugged or broken, it could be that the switch lost electric power, it could be that a large industrial saw is running near the switch and causing interference. The point is there are a limited number of things that could be wrong "physically".

What could be wrong at the network layer? Well it could be a misconfigured IP address or it could be a misconfigured routing protocol, it could even be an access-control list preventing data communication, but it could NOT be an unplugged cable because we already checked that at the physical layer so we don't even need to think about that now.

This approach makes troubleshooting not only easier but also a lot quicker. And fixing problems quickly in a production network is vital.

The 7 layers of the OSI network model are as follows:

The image above also explains what each layer does (or entails). It is good to know what each layer does but it is important especially to know what happens at layers 1 - 4. The truth is that for the CCENT/CCNA exam you won't need to pay much attention to layers above layer 4.

To remember the names of the layers I use the following mnemonic:

All               -Application
People         -Presentation
Seem           -Session
To               -Transport
Need           -Network
Data             -Data Link
Processing    -Physical

Of course there are other mnemonics out there and you can look them up and choose what works best for you.

It is also a good idea to become familiar with the numbers corresponding to each layer. For example a network engineer might say; "that sounds like a layer 3 problem" and you need to know immediately that he's talking about the network layer. The exams also use the layer numbers instead of the layer names quite often so it is really a good idea to get familiar with that.

All-in-all you want to apply this model when considering how data passes throughout a network and when trouble-shooting problems on your network. Take the modular approach and be methodical rather that haphazard and you will be able to identify and correct problems much quicker and with much less effort.



Roger.