From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
WikiProject Computing / Networking (Rated C-class, Low-importance)
WikiProject iconThis article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.
Taskforce icon
This article is supported by Networking task force.

C'mon. What is this article about? Modern dentistry? Space flight? Ice fishing? An encyclopedia article should have at least a short intro that speaks to the generalist. Include the rest of us, please--at least to a degree that we can say, "Oh. I don't care about dentistry."

Arthur 03:13 Mar 8, 2003 (UTC)

I'd love to see some information on new IPv6 capabilities. It'd be cool if there was a wiki-project or similar for networking, like there is for the simpsons, or for theater, etc. One thing that could be added to the IS-IS v. OSPF section is that IS-IS is more scalable, partially because it's easier to extend the backbone. All that's necessary is to add more level-1-2 and level-2 routers. 06:36, 1 May 2007 (UTC)

How scalable is the backbone? I agree that ISIS is easier to build in a flat environment, but you can't just keep adding routers to anything using link state, as even modified Dijkstra algorithms exponentially increase processor load with increases in network links, and linearly in router links. Most large ISP networks have a minimum amount of IGP in their backbones, with the heavy-duty routing carried in BGP, perhaps multiprotocol BGP. Howard C. Berkowitz (talk) 14:05, 5 February 2008 (UTC)

(removed my nonsense comment from ages ago)

ibarrere 05:35 Feb 5th, 2008 (UTC)

May I ask you to cite "official" documentation? In the case of OSPF, an IETF protocol, the only official source is an RFC. In the case of EIGRP, a Cisco proprietary protocol, the documentation is less authoritative in the sense of being third-party.
I've worked with the inventor of the diffusing update algorithm (i.e., the "enhanced distance vector") path computation model used by EIGRP, J.J. Garcia-Luna-Aceves. Dino Farinacci, at Cisco, has been the main EIGRP architect there, and there are authoritative people like Alvano Retara in the protocol scaling group at Cisco in North Carolina. I can't find anything from any of these people saying that protocols for network layer route determination are transport layer. I'm on the IETF mailing lists for OSPF and ISIS, and don't see it mentioned that either are transport layer protocols -- ISIS would have to be network layer if encapsulation is the defining factor, because ISIS travels directly over the data link layer carrying network layer information
At the moment, I'm looking at Alex Zinin's book, Cisco IP Routing. Alex has worked at Cisco, and, unless there's been a recent change, is Routing Area Director for the IETF. The index to the book doesn't contain "transport layer". The ISIS, OSPF, and EIGRP functions do all contain information on the error correction mechanisms built into these protocols, which would meet the OSI definition of transport. Unfortunately for that argument, OSPF is not an OSI protocol, Integrated ISIS is not an OSI protocol, and EIGRP is not an OSI protocol. There's more detail on my userpage.
I spent many years working on pure OSI routing, before I went to the IP side of the Force, and no one ever suggested that ISIS is a transport layer protocol. The OSI Management Framework, ISO document 7498/4, puts routing protocols into the category of network layer management. The OSI Routeing [sic] Framework, ISO/TR 9575, also considers routing protocols network layer. There are more references on my userpage, User:Hcberkowitz#Network Engineering, with more reference citations, including books and RFCs on which I've worked.
I have never seen anything "official" that called a protocol used in finding network layer routes to be called a transport layer protocol. Are BGP and RIP, then, session layer because they, respectively, are encapsulated in TCP and UDP? That's a problem, then, since the IETF doesn't recognize the network layer.
Textbooks -- and I've written some -- are not "official". Please identify the official sources that support your point. Howard C. Berkowitz (talk) 14:02, 5 February 2008 (UTC)

Also: IS-IS is not encapsulated in IP. —Preceding unsigned comment added by (talk) 17:13, 28 February 2010 (UTC)
============================ The OSI model should be reworked on this page ====== 

It looks to me like the whole OSI model on this page should be looked at. I believe that ARP is at layer 2 not layer 3. There also looks to be confusion between routing protocols and routed protocols. I would be happy to edit and correct, if I knew they were not magically going to removed for no reason.

Anyone ever tried searching for this article?[edit]

I tried to search for this article using the Wikipedia search function, and it was quite difficult. I typed in "IS-IS" without the quotes and it came up with completely irrelevent things without even that written in the page such as The Price Is Right game show.

I then tried to search for "Intermediate System to Intermediate System" without the quotes and still it didn't even have that on the first page.

Finally I tried to search for "Intermediate System to Intermediate System routing", without the quotes, and finally it came up with this article as the first result.

I think this article should be boosted in relevency when searching for these terms.

Ayu_Aikousha. —Preceding unsigned comment added by (talk) 01:25, 4 October 2007 (UTC)

Which layer?[edit]

From TFA: "While OSPF is natively built to route IP and is itself a Layer 3 protocol that runs on top of IP"
Mind=blown. IP is a layer 3 protocol (see OSI_Model#Layer_3:_Network_Layer) so how can OSPF run on top of IP if it sits besides it?
Zoef1234 (talk) 12:25, 27 August 2009 (UTC)

IS-IS is not an end-to-end protocol, and is therefore not a transport layer. Like ICMP (which arguably _can_ be end-to-end at times), it is still providing network-layer functionality in terms of supporting routing. —Preceding unsigned comment added by (talk) 04:47, 23 December 2010 (UTC)


Why is there nothing on IS-IS addressing in the article? IS-IS has a somewhat unique addressing scheme and some explanation of it would be nice. (talk) 18:15, 12 October 2010 (UTC)

IS-IS itself doesn't have addressing. The addressing comes from CLNS which is the protocol that IS-IS operates on top of. Like OSPF itself doesn't have any addressing, it runs on top of IP. 2A02:8011:3:0:3898:62AB:544A:CCAA (talk) 10:57, 17 August 2015 (UTC)

Relationship with OSPF[edit]

I think it would be more accurate to say that OSPF was derived from an early draft of IS-IS. Paul Koning (talk) 22:24, 18 December 2012 (UTC)