Talk:Shortest path routing: Difference between revisions

From Citizendium
Jump to navigation Jump to search
imported>Howard C. Berkowitz
imported>David MacQuigg
Line 4: Line 4:


We should not be suggesting that SPF is the Internet-wide routing algorithm, or, indeed, that there is any one algorithm.  SPF is quite effective in networks, or hierarchical subareas of networks.  Even real-world intradomain routing protocols such as [[OSPF]] use  [[distance vector routing]] for their inter-area and external routes. The overall Internet (and large private system) [[Border Gateway Protocol]] uses [[path vector routing]] with a considerable number of policy extensions, the policy so dominating that BGP is not infrequently called a reachability protocol rather than a routing protocol. [[User:Howard C. Berkowitz|Howard C. Berkowitz]] 22:32, 24 December 2009 (UTC)
We should not be suggesting that SPF is the Internet-wide routing algorithm, or, indeed, that there is any one algorithm.  SPF is quite effective in networks, or hierarchical subareas of networks.  Even real-world intradomain routing protocols such as [[OSPF]] use  [[distance vector routing]] for their inter-area and external routes. The overall Internet (and large private system) [[Border Gateway Protocol]] uses [[path vector routing]] with a considerable number of policy extensions, the policy so dominating that BGP is not infrequently called a reachability protocol rather than a routing protocol. [[User:Howard C. Berkowitz|Howard C. Berkowitz]] 22:32, 24 December 2009 (UTC)
:Good points.  I can see there are some unintended suggestions in the second paragraph of the intro.  How about "This article will explain a basic routing algorithm [1] commonly used in routing protocols for small to mid-sized networks."  I'm still a little puzzled by what this means in real numbers.  Peterson & Davie (p.267) say "fewer than a hundred nodes".  The Python program in this article will compute a routing table of 3000 nodes in 1.2 seconds.  In C, this would be more like 12 msec.  Surely this is fast enough for a network where link costs change over a period of hours. --[[User:David MacQuigg|David MacQuigg]] 20:57, 25 December 2009 (UTC)

Revision as of 14:57, 25 December 2009

This article is developed but not approved.
Main Article
Discussion
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
Catalogs [?]
 
To learn how to update the categories for this article, see here. To update categories, edit the metadata template.
 Definition finding the shortest path through a network [d] [e]
Checklist and Archives
 Workgroup category Computers [Please add or review categories]
 Talk Archive none  English language variant American English

SPF runs in small routing domains

We should not be suggesting that SPF is the Internet-wide routing algorithm, or, indeed, that there is any one algorithm. SPF is quite effective in networks, or hierarchical subareas of networks. Even real-world intradomain routing protocols such as OSPF use distance vector routing for their inter-area and external routes. The overall Internet (and large private system) Border Gateway Protocol uses path vector routing with a considerable number of policy extensions, the policy so dominating that BGP is not infrequently called a reachability protocol rather than a routing protocol. Howard C. Berkowitz 22:32, 24 December 2009 (UTC)

Good points. I can see there are some unintended suggestions in the second paragraph of the intro. How about "This article will explain a basic routing algorithm [1] commonly used in routing protocols for small to mid-sized networks." I'm still a little puzzled by what this means in real numbers. Peterson & Davie (p.267) say "fewer than a hundred nodes". The Python program in this article will compute a routing table of 3000 nodes in 1.2 seconds. In C, this would be more like 12 msec. Surely this is fast enough for a network where link costs change over a period of hours. --David MacQuigg 20:57, 25 December 2009 (UTC)