Talk:Routing

From Wikipedia, the free encyclopedia

Contents

[edit] Page contents

This page fairly desperately needs to be split up into:

  • what a next hop is (via an interface or gateway),
  • how routing works by selecting the most specific route to a destination (using longest prefix match),
  • what happens when there are two identical routes to the same destination but using different routing protocols (route preference),
  • what a default route is,
  • classful versus CIDR routes (and why the latter became necessary),
  • etc.
This is basic stuff!
What this page contains is how entries are added and removed from the routing table. It should go on a separate page about routing algorithms/protocols as you mention above. Mordomo 06:28, 20 February 2006 (UTC)

[edit] EIGRP is not a link-state protocol

I'm getting really tired of seeing the Cisco marketing balderdash about EIGRP being a "hybrid" of link-state routing and destination-vector routing spammed across Wikipedia, and even more tired of seeing repeatedly inserted after I keep removing it. I'm therefore going to spam this across every Talk: page where I see this claim, and a shorter note to the effect that EIGRP has no link-state stuff at all, in the articles.

Nothing could be further from the truth than the claim that EIGRP has any link-state aspects.

EIGRP is simply a multi-metric, event-driven, destination-vector routing protocol. Neither the "multi-metric" part nor the "event-driven" part has anything to do with link-state.

Link-state protocols have the following characteristics:

  • they distribute topology maps, not routing tables
  • nodes run a shortest-path algorithm such as Dijkstra over the map to produce the routing table

EIGRP does neither.

Clearly, one can design link-state protocols to be either event-driven, or not; all done to date (from the original "new" ARPANet routing algorithm) have been so, but that's purely a design decision. Event-driven or not-event-drive is a completely separate design axis.

Now stop adding this bogus nonsense! Noel (talk) 04:57, 24 Dec 2004 (UTC)

[edit] "Hybrid" protocols

I removed the following text from the page:

There is also a third method called hybrid: Hybrid protocols are a combination of link-state and distance-vector routing protocols. Hybrid protocols have rapid convergence (like link-state protocols) but use much less memory and processor power than link-state protocols. Hybrid protocols use distance-vectors for more accurate metrics and to determine the best path to destination.

because most of it's untrue. The only true MD/DV hybrid (it wasn't even link-state, but rather Map-Distribution, a larger class that includes link-state) ever even proposed (that I know of) was the "Unified" design of Rehkter and Estrin, circa 1988 or so (Deborah Estrin, Yakov Rekhter and Steve Hotz, "A Unified Approach to Inter-Domain Routing", RFC 1322) but it did not have the characteristics of "rapid convergence ... but use much less memory and processor power than link-stat

[edit] routing metric

It is not clear from the article if a higher value means a route is more or less likely to be chosen. It also doesn't make clear what happens if two routes have the same metric.

[edit] Value of ON content and quality of reference

The content added from the ON reference remains in this article, but the reference has been removed. This action is disputed and a conversation is ongoing here. Uriah923 06:18, 2 September 2005 (UTC)

[edit] Replace external link?

The following was recently removed from the article for being spam unless a consensus could be reached here that it is not: "See [(see this page's history for the link) this article] for a good example of a distance vector and link state algorithm application."

I think the link should remain as it is quality and valuable content. Uriah923 16:51, 12 September 2005 (UTC)

I'll say pretty much the same thing here as I did at Talk:ITunes. This is just one out of a SEO campaign by the ON people to get links to that site from Wikipedia. (Just as we don't need the above link since it can be seen in the diff where it was removed) The same thing was done with multiple Wikipedia articles until people cried foul, and the consensus was that ON articles do not make the type of quality references Wikipedia needs. That said, if a consensus forms here (with a reasonable minimum of 5-6 people involved) that the article is valuable enough to justify an external link despite the linkspam implications, I certainly wouldn't stand in the way. That said, our article would be much better off using the actual RFC's and highly regarded networking textbooks as references than linking to a website with no inherent credibility. Wikipedia:Verifiability is the goal, not seeing how many links we can get to ON. - Taxman Talk 18:33, September 12, 2005 (UTC)
We are not discussing the use of the site as a reference, but leaving the aforementioned sentence. My contention is that it should be replaced. As far as I can surmise, you are against it's addition purely due to SEO suspicions. Can we have some people evaluate this on a purely content basis? Uriah923 19:59, 12 September 2005 (UTC)

Remove the darn ON link. It is not verifiable or reliable. Agree with Taxman that RFCs and textbooks (of which I have a slew) make better references. Zora 21:09, 12 September 2005 (UTC)

It's too bad there isn't an ON link to remove or that we aren't talking about references or else your comment might make sense. Uriah923 22:09, 12 September 2005 (UTC)

[edit] Nullrouting

Please add a link to Nullroute if it belongs in this article or any other routing related article.

[edit] Ether

Please help clarify the entry in the "aether" disambiguation page: "in internet routing, the term ether is associated with hosts" What the heck is it? (Previously "hosts" was a redirect to "hosts file") `'mikka (t) 17:38, 17 July 2006 (UTC)

What procesess/applications instigates routing algorithms in unix and windows systems and what are the files involved??

Retrieved from "http://en.wikipedia.org../../../r/o/u/Talk%7ERouting_087f.html"