Talk:TCP/IP model

From Wikipedia, the free encyclopedia

	This article is part of WikiProject Computer networking, an attempt to build a comprehensive and detailed guide to Computer networking on Wikipedia. If you would like to participate, you can edit the article attached to this page, or visit the project page, where you can join the project and/or contribute to the discussion.	Computer networking Portal
Start	This article has been rated as start-Class on the assessment scale.
High	This article is on a subject of high priority within Computer networking for inclusion in Wikipedia 1.0.
Computer networking To-do: edit  Computer networking Collaboration of the week: Wikipedia:Computer networking Collaboration of the week/current Template:Cnbox contents

Contents 1 General Cleanup Needed 1.1 Howard C. Berkowitz - 14:59, 17 December 2007 (UTC) 1.2 SP - December 17th, 19:45:00 (EST)(GMT-5) 2 DoD model 3 Requested move 3.1 Survey 3.2 Discussion 4 Time for another move? 5 easier fit for real-world protocols? 6 layers all wrong? 7 Five-layer model 8 To clear up some confusion and note a mistake in the content of the TCP/IP model page 9 Broken link

[edit] General Cleanup Needed

I added two tags tonight: copy-editing, and confusing-or-unclear

Additionally:

- Structure - the article may need structural clean up as most of the intro should be it's own section, or may be repeated in later sections.

- Slant - "IETF", the presiding technical body over TCP/IP is referenced 12ish times. 9 of them are in a negative context. This worries me.

- Expert-Needed - Some of the facts seem fuzzy, and discussion-page sub-titles seem to bring them into question. Perhaps an expert should be consulted.

Overall, it's clear this article has been pieced together by engineers, not English-majors. There are plenty of 'minor' grammar, tone, style problems but they quickly add up. Some bias slant is detectable. An industry-expert and a professional editor could probably make short work of this heavy-cleanup. (alas, I am neither..)

On a broader view - sections of the article seem to debate view-points on the subject matter rather than concisely and authoritatively presenting the subject. Also, article spends alot of time on what TCP/IP is NOT, and what the IETF does NOT allow when it mentions them at all. I did not make it to the part where the article explains what TCP/IP is.

For example the entire introduction debates whether TCP/IP should be taught as a layered model or not, and says nothing about TCP/IP! (1st who-cares, 2nd why in the intro?)

I'm overtired and overstepping the bounds of my knowledge but the intro could go something like this. In fact it probably needs nothing more - anything else should be in a later secton. (Note: this is for example, I can't personally vouch for the factual-accuracy here.)

The TCP/IP Model is sometimes called the Internet Reference Model, the DoD Model (DoD, Department of Defense), or ARPANET Reference Model. Written in the 1970s by DARPA (an arm of the American DOD), it is a high-level design document for computer network protocols. It laid the foundation for ARPANET, the worlds first Wide-Area-Network, which eventually grew up to be the modern day internet. While technology has advanced considerbly, the TCP/IP model still lays the foundational structure for the modern internet.

Simply put, TCP/IP defines how two or more computers should communicate over a network. This includes how such communications should be packaged, addressed, shipped, routed and delivered to the correct destination. It creates the framework for more specific standards and protocols, which define specific types of communication between computers. These protocals are also said to fall under the TCP/IP Model.

The model is generally taught as being a four-layer model; or five-layers if you include the bottom physical layer. This concept is primarily based on the seven-layer OSI Reference Model, written long after the TCP/IP specs and is not offically recognized. Regardless, it makes a good analogy for how TCP/IP works and the comparison is common.

The TCP/IP model and related protocals are currently maintained by the Internet Engineering Task Force (IETF). IETF is a "large open international community of network designers, operators, vendors, and researchers concerned with the evolution of the Internet architecture and the smooth operation of the Internet." http://www.ietf.org/overview.html

--64.253.28.131 (talk) 09:49, 17 December 2007 (UTC) SP

OK -- I've had a go at copy-editing the above introduction, which I think is much better than what the article has at the moment. I have no in-depth knowledge of the subject, so apologies in advance for any mistakes of fact.

On a general (pessimistic!) point, I think this article is a prime example of where Wikipedia fails at the moment: an important technical article that is baffling and dominated by secondary technical issues.

The TCP/IP Model is a specification for computer network protocols created in the 1970s by DARPA, an agency of the United States Department of Defense. It laid the foundations for ARPANET, which was the world's first wide area network and a predecessor of the Internet. The TCP/IP Model is is sometimes called the Internet Reference Model, the DoD Model (DoD stands for Department of Defense) or the ARPANET Reference Model.

TCP/IP defines a set of rules to enable computers to communicate over a network, specifying how data should be packaged, addressed, shipped, routed and delivered to the right destination. The specification defines protocols for different types of communication between computers and provides a framework for more detailed standards.

TCP/IP is generally described as having four 'layers', or five if you include the bottom physical layer. The layer view of TCP/IP is based on the seven-layer OSI Reference Model written long after the original TCP/IP specifications, and is not officially recognized. Regardless, it makes a good analogy for how TCP/IP works and comparison of the models is common.

The TCP/IP Model and related protocols are currently maintained by the Internet Engineering Task Force (IETF).

Sam Dutton (talk) 17:42, 4 March 2008 (UTC)

That's good, I like it. It's much more concise and readable than the current intro. Indeterminate (talk) 20:16, 4 March 2008 (UTC)

I've tweaked it a bit. Sam Dutton (talk) 11:39, 5 March 2008 (UTC)

OK: I've changed the introductory text. It would be great if someone with more TCP/IP knowledge could check it. Sam Dutton (talk) 21:28, 11 March 2008 (UTC)

[edit] Howard C. Berkowitz - 14:59, 17 December 2007 (UTC)

On a broader view - sections of the article seem to debate view-points on the subject matter rather than concisely and authoritatively presenting the subject. Also, article spends alot of time on what TCP/IP is NOT, and what the IETF does NOT allow when it mentions them at all. I did not make it to the part where the article explains what TCP/IP is.

I agree completely that there is too much about how TCP/IP is not OSI, but that is due to the history of the article; it has improved significantly since I first saw it. Unfortunately, many beginners to networking get an introduction from an obsolete view of OSI. For months, I've been removing references that marginally refer to OSI as ISO (i.e., not basic tutorials) understands it, and that also try to force-fit TCP/IP into OSI.

For example the entire introduction debates whether TCP/IP should be taught as a layered model or not, and says nothing about TCP/IP! (1st who-cares, 2nd why in the intro?)

I agree completely. The IETF, which, as you point out, is the authority, has architectural documents that argue against strict layering. Just in the last few days, however, there's been an argument about the layering of routing protocols, and, when I offered both authoritative documents and personal experience (in writing some of those documents), what came back was an argument.

I'm ... overstepping the bounds of my knowledge but the intro could go something like this. In fact it probably needs nothing more - anything else should be in a later secton. (Note: this is for example, I can't personally vouch for the factual-accuracy here.)

Believe me, I feel as frustrated, and, with no false modesty, am speaking within the bounds of my knowledge. There's just as much misinformation, due to marketing usages, of such things as "switch" rather than "bridge", "router", or a combination. There were a few cases where an editor insisted a device worked a given way and that it was an (undefined) "switch" rather than a router inside, because the product name, for historical and marketing reasons, was switch. In one of the cases (Nortel 8600) I was among the designers, and for some Cisco devices, I was involved in the IETF work, the Cisco product testing, and reasonably accurate education on it.

There's no simple answer. Personally, I have gotten tired of the never-ending arguments based on incorrect information, and am considering giving up editing networking articles; it's too frustrating.

Howard C. Berkowitz (talk) 14:59, 17 December 2007 (UTC)

[edit] SP - December 17th, 19:45:00 (EST)(GMT-5)

There's no simple answer. Personally, I have gotten tired of the never-ending arguments based on incorrect information, and am considering giving up editing networking articles; it's too frustrating.

Wow! NO IDEA I was stepping into a firestorm. But maybe glad I did!

I don't understand what there is to debate here? If we already have a resident expert who helped design this stuff - only another equally ranking EXPERT would be qualified to make real technical arguments. Howard, you should not allow yourself to be bogged down in such meaningless arguments in life. It's not worth it. If you know your stuff, show it, and move on.

I'm tempted to say just write the damn article and email it to me. I'll edit and revise it. If someone wants to argue, we will appeal to wikipedia, and it will be out of our hands.

I'm not an expert in this topic but I have a sufficient technical background and can be a decent writer when I try. I've taken a sudden interest in this subject matter, and the chance to work with someone who was actually there - it would be cool.

These topics should not be up for debate.

One-Day-Link (Contact-Info)http://codeshare.ulatu.com/index.cfm/view/review/shareID/dec949ec

--64.253.26.36 (talk) 03:21, 18 December 2007 (UTC) SP

Previously this article was called DoD model.

[edit] DoD model

Pardon me, but what has this to do with the American Department of Defense? I am confused and angry for unspecified reasons. —The preceding unsigned comment was added by 84.227.10.12 (talk • contribs) .

Perhaps because it was developed by DARPA, which is an agency of the Department of Defence? Bryan 00:21, 8 February 2006 (UTC)

[edit] Time for another move?

The opening paragraph now speaks of the TCP/IP reference model rather than the TCP/IP model - the TCP/IP reference model page is a redirect to TCP/IP model, but, with that change, it sounds as if the primary page should be called "TCP/IP reference model" with "TCP/IP model" being a redirect to it. Unfortunately, there's no "swap" operation that lets any user exchange a page with a redirect to that page, so, to make that change, we'd have to request another move by an administrator; should we do so? Guy Harris 18:38, 21 November 2006 (UTC)

Google gives more hits on "TCP/IP model", so I reverted my change of the opening paragraph. Instead I added "Internet reference model" as an alternative term. That solves the problem. A move is not necessary. Mange01 21:34, 21 November 2006 (UTC)

[edit] easier fit for real-world protocols?

How is this model an easier fit? Given that the next paragraph tells us there is no single version of the model with different numbers of layers, I fail to see how this makes it an easier fit Markb 12:59, 5 February 2007 (UTC)

[edit] layers all wrong?

I may be missing something, but how can RIP, which layers over UDP, be lower in the network stack than UDP? Also IPSec, ICMP and IGMP are over IP. ESP is a sibling of TCP/UDP etc. the only protocols listed there that should be at layer 3 are IP, ARP and RARP (which isn't even strictly it's own protocol, but an extension to ARP), e.g only those protocols that have a frame type number in the Ethernet header.. Layer 4 should be protocols that have their own protocol number in the IP header e.g.: TCP(6), UDP(17), ICMP(1), IGMP, ESP, IPIP, GRE.

L2TP is over IP, and PPTP uses a TCP connection (port 1720) for control channel and GRE for data tunnel, so most definitely isn't at the Data Link Layer below IP. —Preceding unsigned comment added by 125.237.240.118 (talk • contribs) 11:23, June 21, 2007

With the caveat that I have never seen on an IETF mailing list, or heard in a serious discussion at an IETF meeting, any particular concern about the layer in which a protocol resides, it is the function of a protocol, not how it is encapsulated, which defines the layer at which a protocol resides. Even in formal OSI protocol work, routing protocols are assigned to the network layer by the Management Framework and the OSI Routeing Framework, as layer management protocols.

L2TP, which most people would say is a layer 2 protocol if they had to assign it, runs over UDP over IP.

Layering is a guide that doesn't always work with real-protocols. I'm mystified, incidentally, why the graphic shows five rather than four layers for this model. Howard C. Berkowitz 21:39, 2 July 2007 (UTC)

Hey, I second that, I used this information in an Exam and I failed because of it... I didn't consider RIP as an Application Layer protocol. You should check out what the standards say and maybe look a little over the Cisco CCNA's! —Preceding unsigned comment added by MieZu (talk • contribs) 15:29, 9 June 2008 (UTC)

[edit] Five-layer model

I quite seriously propose that the five-layer stack drawing, which is not authoritative by any IETF document, be replaced with a four-layer stack consistent with RFC 1122. As far as I can tell, the basis for the five layer argument are only secondary sources, such as Stallings' book.

As long as I am being bold, I'd like an agreement to put strong words in the introduction to this article that it is not OSI-compliant, there is strong historical reason that the designers of this stack did not intend it to be OSI comppliant, and that people experienced in using and teaching this stack find that one of the chief barriers to understanding the most widely used stack in the world is that they are trying to force it into an obsolete OSI model. Howard C. Berkowitz 22:29, 13 October 2007 (UTC)

I agree with you that a four-layer stack makes more sense and would be more appropriate... however, I think it's important to keep an article (or part of an article) on the five-layer model, if only because it's so well-known. If you want to create an article for the four-layer model (or maybe better, just expand this article), I think it would be an excellent idea.

But it sounds like you might just be interested in replacing that awkward TCP/IP Model template, which is actually maintained over in the Template area: Template:IPstack. It looks like they've been arguing on the Talk page about the number of layers to include for quite a while. :)

Either way, I approve of both your proposals. Indeterminate 07:18, 16 November 2007 (UTC)

Thank you. I'll try to clone this over at the Template area, but I've become frustrated about making any headway over here, because of loud and unsupported arguments that there is a five-layer model. I'm puzzled by the need to keep an article on the "five-layer model", which does not exist in any IETF documentation, but explicitly defines a four-layer model (RFC1122), and explicitly speaks against strict layering in several statements of architectural principles. ISO does not ever define a five-layer model; ISO started with a seven-layer model in ISO 7498, and then added sublayers in several documents. It does not exist in any IEEE document, which deal only with the bottom two layers and sublayers. These, I believe, constitute the primary sources on networking architecture and layering.

Further, it does not exist in any training material from Cisco, Juniper, or Nortel, and, as far as I know, Microsoft. The major vendors arguably are the chief secondary sources of de facto rather than de jure concepts of layering.

It exists, as far as I can tell, in some textbooks, and in Wikipedia. I've published textbooks myself, and I have asked a number of colleagues who are also networking book authors, and they never used the concept. If one makes use of the sublayering of IEEE or ISO, even ignoring the reorganization of the upper layers by ISO, one might get:

• 9 layers: OSI basic model plus IEEE sublayering of LLC/MAC and (two layers, but by different names in different documents), medium independent/medium dependent, or PLS/AUI in the original Ethernet documents.
• 11 layers: OSI plus the additional two layers of the Internal Organization of the Network Layer, plus the additional layers in IEEE 802.1 mentioned just above.
• 7 layers, IETF plus IEEE, taking the three layers from internetwork on up, and then two two-sublayer models below it.
• 9 layers, using the three sublayers of the ISO IONL, and the two sublayers of IETF in the bottom two layers, plus IETF end-to-end and application.

In the real world, however, all of these tend to break down when considering protocols that use recursion to create tunnels, such as L2TP, IP in IP, IPSec, MPLS (with arbitrary numbers of stacked labels), 802.1q-in-q, ATM LAN Emulation, etc.

I appreciate your comment, and will try to bring this to the template talk page. Unfortunately, I don't know how to create a revised template and propagate it, and, unless there is first some consensus that no authoritative reference uses five layers, it doesn't seem worth the effort if some people are going to keep citing a textbook or two that, as far as I can tell, is the only source. Howard C. Berkowitz 09:14, 16 November 2007 (UTC)

As a networking instructor, I sure would appreciate switching to the 4-layer model. Students get confused when looking up TCP/IP and finding a hybrid 5-layer model that is referenced nowhere in their materials. —Preceding unsigned comment added by 83.233.183.35 (talk) 00:36, 15 January 2008 (UTC)

[edit] To clear up some confusion and note a mistake in the content of the TCP/IP model page

I'm going to be brief. There are only 4 layers in the TCP/IP model, you have them all right i think besides the network and data link part. Those two are in the OSI 7 layer model. The network layer is called the Internet layer in TCP/IP and the physical and data link layers are fused together in the TCP/IP model and are just called the physical layer. BTW the DoD has everything to do with the TCP/IP model...since they you know...created it... —Preceding unsigned comment added by 66.56.90.29 (talk) 21:34, 7 March 2008 (UTC)

[edit] Broken link

The URL for reference 2: "Rethinking the design of the Internet: The end to end arguments vs. the brave new world,D. Clark & M. Blumenthal, August 2000" http://www.tprc.org/abstracts00/rethinking.pdf is a broken link ("Page not found") There appears to be another copy of this document at the following URL: http://www.ecse.rpi.edu/Homepages/shivkuma/teaching/sp2001/readings/Rethinking_2001.pdf

Can someone confirm that this is correct and fix the link? Thanks. —Preceding unsigned comment added by 192.28.2.6 (talk) 21:02, 22 May 2008 (UTC)

The link was fixed by the time I got to it. Indeterminate (talk) 22:17, 9 June 2008 (UTC)