ISO/OSI is a neatly separated model mostly used on theory.
In practice, actual network stacks are often modeled after a simpler model that is called TCP/IP. Which despite the name is not actually TCP specific.
Here’s the general description and correspondence to ISO/OSI:
Host to network / network access layer: it’s mostly the nic and nic driver. It’s sometimes numbered as 0 because some don’t consider it part of the TCP/IP stack, but simply the nic driver. Corresponds to:
Physical
Datalink
Network layer: Corresponds to:
3. Network
Transport layer: Corresponds to:
4. Transport
Application layer: everything that’s part of the application and not the network stack. Corresponds to:
5. Session
6. Presentation
7. Application
Or, you can just not care about how the actual software stack is separated, and continue to use the most complete model, knowing that everyone will understand what you when you say “layer 2/3/4” anyway.
Plus, some could say that the TCP/IP model is equally unfit because the Linux network subsystem doesn’t care about layers.
Edit: I hope the formatting of that table isn’t broken on your client, because it is on mine
The IP suite is not nearly as neatly layered as OSI was and the OSI model doesn’t neatly fit the IP suite since it wasn’t actually designed for IP at all. In the IP suite layers 5 and 6 basically don’t exist in the OSI sense, TCP handles things that are part of both layers 4 and 5 in the model despite being a single protocol, etc. The OSI model is often considered obselete as it just doesn’t actually fit the IP world all that well but it’s been around so long and does have uses in certain situations that it tends to stick around.
ISO is a theoretical model on how things work, suitable for teaching and for reference. In reality, there are only 4 layers and not the finely chiseled concern-separating layers at the bottom.
IP is the internet layer which is layer 2 and TCP is the transport layer which is layer 3.
Why split physical and data link when they are so closely related? You can’t use vpi/vci on an ethernet port, you can’t use MAC addresses on a frame relay port. Bundle that shit together.
Presentation, application, and session are all dealt with by the application anyway so why bother splitting them out from a network point of view?
Why split physical and data link when they are so closely related?
You can run Ethernet on any medium that has the capacity to transmit digital signals. It can be copper, optical, over-air laser, radio, on top of an analog carrier wave (ASK, FSK, PSK). The Ethernet traffic can be completely independent from the physical medium by using encapsulation (L2TP or any other protocol that encapsulates Layer-2). It can be pigeons carrying printouts of the Ethernet frames, scanned and reassembled at the destination. The same can be said about most Layer-2 protocols.
As long as the proper interfaces are present, the physical layer is completely transparent to the data link layer.
(edit) I should point out that Ethernet, specifically, transmits extra data before and after the frame (the preamble and inter-packet gap) that are used to configure the Rx circuit for reception, but the Layer-2 frame will be identical regardless of the medium.
Why split physical and data link when they are so closely related?
100 mb Ethernet can be run over cat 5e, Cat 6, etc. You can even have Ethernet frames running over coax. (Not just the original standard which was coax but there are modern bridges for using cable TV coax in your house to reach rooms where you can’t pull cat5.)
Easier way is to remember that the ISO model is defunked and you should use the TCP/IP model.
TCP/IP model is Limited and incomplete.
Fibrechannel, infiniband, and many RDMA protocols, just to make a few, don’t fit in the TCP/IP model, while the OSI model fits them just fine.
How don’t they fit?
IP is layer 3 and TCP is layer 4.
Please elaborate your nonsensical comment.
ISO/OSI is a neatly separated model mostly used on theory.
In practice, actual network stacks are often modeled after a simpler model that is called TCP/IP. Which despite the name is not actually TCP specific.
Here’s the general description and correspondence to ISO/OSI:
Or, you can just not care about how the actual software stack is separated, and continue to use the most complete model, knowing that everyone will understand what you when you say “layer 2/3/4” anyway.
Plus, some could say that the TCP/IP model is equally unfit because the Linux network subsystem doesn’t care about layers.
Edit: I hope the formatting of that table isn’t broken on your client, because it is on mine
The IP suite is not nearly as neatly layered as OSI was and the OSI model doesn’t neatly fit the IP suite since it wasn’t actually designed for IP at all. In the IP suite layers 5 and 6 basically don’t exist in the OSI sense, TCP handles things that are part of both layers 4 and 5 in the model despite being a single protocol, etc. The OSI model is often considered obselete as it just doesn’t actually fit the IP world all that well but it’s been around so long and does have uses in certain situations that it tends to stick around.
ISO is a theoretical model on how things work, suitable for teaching and for reference. In reality, there are only 4 layers and not the finely chiseled concern-separating layers at the bottom.
IP is the internet layer which is layer 2 and TCP is the transport layer which is layer 3.
Why split physical and data link when they are so closely related? You can’t use vpi/vci on an ethernet port, you can’t use MAC addresses on a frame relay port. Bundle that shit together.
Presentation, application, and session are all dealt with by the application anyway so why bother splitting them out from a network point of view?
You can run Ethernet on any medium that has the capacity to transmit digital signals. It can be copper, optical, over-air laser, radio, on top of an analog carrier wave (ASK, FSK, PSK). The Ethernet traffic can be completely independent from the physical medium by using encapsulation (L2TP or any other protocol that encapsulates Layer-2). It can be pigeons carrying printouts of the Ethernet frames, scanned and reassembled at the destination. The same can be said about most Layer-2 protocols.
As long as the proper interfaces are present, the physical layer is completely transparent to the data link layer.
(edit) I should point out that Ethernet, specifically, transmits extra data before and after the frame (the preamble and inter-packet gap) that are used to configure the Rx circuit for reception, but the Layer-2 frame will be identical regardless of the medium.
100 mb Ethernet can be run over cat 5e, Cat 6, etc. You can even have Ethernet frames running over coax. (Not just the original standard which was coax but there are modern bridges for using cable TV coax in your house to reach rooms where you can’t pull cat5.)