OSI Model

I studied the OSI model, but nobody could ever tell me why I need to know it. The answer I always got was "you definitely need to know the OSI model" but when I ask why I always get "you just do".

Can anyone tell me why I need to know it and how I would apply the knowledge? Perhaps a lab exercise or something??

Thanks, CJ

James Knott prattled ceaslessly in news: snipped-for-privacy@rogers.com:

TCP/IP is a protocol suite. The OSI model is a general model for networking. TCP/IP follows the OSI and DOD model and the various protocols in the suite fit into the model at different layers.

James Knott prattled ceaslessly in news: snipped-for-privacy@rogers.com:

Once again, OSI is not a protocol stack. It is a general model that was actually followed when developing TCP/IP as well as other protocol stacks.

You need to know it just like other kinds of terminology: to understand people and publications which use it to illustrate something that they want to teach you.

Definitely not. The OSI model has no relevance in _practise_.

best regards Patrick

It's just a method of learning the general layers of the protocol stack. It's also a stack that was intended to be used, but was soon surpassed by TCP/IP.

The issue is that the people who create all of this networking stuff use this model to simplify what would otherwise be a more difficult problem.

This has a couple of consequences:-

If you are designing or troubleshooting a network then it helps to be able to think about the kit in the same way as the designers and implementors do.

In a purer application of the model, the most critical phase of troubleshooting for me is to determine what layer the problem resides in.

e.g. 1. If the link light is not on it is an L1 problem and we do not (yet) need to worry about possible IP addressing issues (L3).

e.g. 2. If we are getting ARP entries in our arp cache for the target then L2 must be working since we are exchanging frames. There is no point in worrying about STP any further.

In many cases this is just _obvious_ however in more subtle cases or when learning about networking it can help a great deal to have a formal structure to hang on to.

The KEY idea is that the respective layers at each end of a communications link talk DIRECTLY to each other.

If L3 at end A can talk to L3 at end B then L1 and L2 and L3 must be working correctly at both ends and at all points between.

Finally, it is a MODEL and for example, TCP/IP is quite a close fit to the OSI model but it seems to be generally accepted that it is a 4 layer system with layes 5, 6, 7 of the OSI model not being applicable.

PS The critical thing is not knowing the names of the 7 layers (which I don't, unless you need to pass tests) but having the concept in your head that TCP/IP has been constructed using a layered model.

Actually, at one time, it was intended to be the "standard", but instead, TCP/IP became the defacto standard.

So you can read all those articles in Datamation and ComputerWorld in the 80s that argued whether or not the OSI model should be a requirement or not and whether or not IBM's SNA was OSI compliant and .....

It got rather tedious after a while.

ISTR that LDAP is a version of X.400, another standard from the same period.

Remember to keep OSI (Open System Interconnect) and ISO (International Standards Org). The fact that the ISO standards were created by ISO can be confusing.)

The 7 layer model OSI is important and very useful when you are trying to understand different protocols, ie IP and SS7, even in one or both are not OSI-based, and figure out how to make them work together. It's a common framework when you are trying to explain how some component fits with other components. Try explaining how IP runs over a broadcast medium (ethernet) and also a serial connection without refering to layers.

For a few years the US gov't had a policy that said as of some date all equipment purchased had to conform to relevant OSI standards and there was lots of lip service but departments started to buy TCP-based systems anyway. At some point they came out with GOSSIP (Government Open Systems .something... Plan) to tray to make something of themess. Google for it.

Padlipsky's _The Elements of Networking Style_ is a great book about how IP and ISO competed in the 80's. Every communications book by Tannenbaum I've seen has covered the 7 layer model well.

??? The OSI model was first proposed in 1977 published in preliminary form in March 1978, with a second preliminary release in June, 1979, and adoption as a standard coming later. TCP/IP was first defined in 1973 as TCP, and split out into TCP and IP as separate components in TCP/IP Revision 4, released in January, 1978.

So it's difficult to attribute any part of the development of TCP/IP to the influence of the OSI model. It could be that the 1979 or 1980 or later specification revisions may reference the OSI model in some manner but that is not the same as using the model when developing the protocol.

>

James Knott prattled ceaslessly in news: snipped-for-privacy@rogers.com:

The OSI model is not a protocol suite. TCP/IP is a protocol stack based on the OSI or DOD model. TCP/IP did not replace the OSI model, it followed it. The OSI model is a very general specification for how to build a protocol stack.

No argument here. Some parts of those sets of standards even survive today, for example the IS-IS routing protocol, which is used as an interiour routing protocol in lots of contemporary IP networks.

best regards Patrick

Since TCP/IP preceded the OSI model it's difficult to see how it could be based on that model.

As for the DOD model, since TCP/IP was developed as part of a DOD advanced research project, one would expect the DOD model to be based on TCP/IP.

"CJ" prattled ceaslessly in news:12W6e.726$ snipped-for-privacy@monger.newsread.com:

No, it is useful to know if you want to understand how networks work. It is also necessary if you are contemplating any type of certification in networking. Those layers will come up over and over again.

Until some guy in a job interview asks you to explain your understanding of the OSI model.

IP and parts of the ISO _standard_ were co-developed. The problem is that RFC standards process emphesised practical working functionality designed by the programers that wrote the code as they went and the OSI standard was written by teco engineers who included every function they could think of without any thought of was it was possibe to code. It took years for products based on OSI mode to be useful.

I understand how OSI works in general. But I can know alot of information about alot of things....doesn't mean its of any use to me.

From what I am reading knowing the OSI is a "nice to know" thing, but knowing it, unless I am a protocol developer or something, is useless to me.

Sound about right?

At one time, OSI was intended to be the common networking method, but didn't make it. It was at that point that it was declared a "model".

From

"One interesting aspect of the history of the OSI Reference Model is that the original objective was not to create a model primarily for educational purposes-even though many people today think that this was the case. The OSI Reference Model was intended to serve as the foundation for the establishment of a widely-adopted suite of protocols that would be used by international internetworks-basically, what the Internet became. This was called, unsurprisingly, the OSI Protocol Suite.

However, things didn't quite work out as planned. The rise in popularity of the Internet and its TCP/IP protocols met the OSI suite head on, and in a nutshell, TCP/IP won. Some of the OSI protocols were implemented, but as a whole, the OSI protocols lost out to TCP/IP when the Internet started to grow."