Quote:
|
Originally Posted by erol
With dialup there is 'contention' issue in terms
|
OK first off I'll get the appologies out of the way
appologies for replying to my own post.
appologies if this is off topic and should go somewhere else (please move it as appropriate)
appologies if this is of no interest to anyone.
I was thinking some more about ways to explain the idea that higher speed connections generally can support higher contention ratio. I think I have a way that helps make the point clearer (and some other about contention) - that uses my favourite tool - an analogy (analogy warning - analogy are not exact comparisions but ones designed to highlight a particular aspect only. Outside of that aspect they usualy break down horribly).
So in the analogy imagine that
The shared resource of a contented data pipe (where ever that contetion might be - first/last mile, middle mile or external connectivity) is a public toilet.
A users connection speed relates to how quickly they can 'evacuate' themselves. A 1mbs users takes half the time to 'evacuate' as a 512kbs user.
Contention then becomes when you go to use the toilet and it is already in use and you have to wait.
Once you do this then the point (that the more you increase a users speed the more contention you can stand) becomes (I hope) much clearer. Because people spend less time doing their doings (as their connection/ evacuation speed increases) they are in and out of the toilet quicker and thus the chance that when you go to use the toilet it is already occupied is lower. Thus you can increase the number of people sharing the toilet when you increase their evacuation speed. This of course has the implicit assumption that just because people can now evacuate quicker they do not decide to evacuate more.
This analogy (imo) is also useful a looking at some other contention ideas. Like the idea that contending more people on a larger pipe has less impact on users than less people on a small pipe at the same contention ratio.
If you imagine a public toilet with a single bowl being shared by say 20 people - giving a TCR (toilet contention ratio) of 20:1
and then imagine a public toilet with 10 bowls being shared by say 200 people - giving the same TCR of 20:1
In the first example a single indivdual that takes ages and ages to evacuate is all it takes to cause severe blockage (possibly not the right term given the anaology used). In the second example it would take 20 such 'long time' evacuators - all evacuating at the same time, which is a lot less likely than there being one. Thus in general terms the bigger the shared pipe, shared by more users is, at constant contention ratios the less impact (relative to small pipe with less users at same contetion)
The analogy can also be used to look at the CAP and heavy / abusive users as well.
In this analogy data volumes downloaded (or uploaded) relate to amount evacuated.
NTL currently define toliet abuse as being based on the amount people evacuate. However there may be a user that whilst they evacuate 10 or even 100 times more than the 'average' user they always do so between the hours of midnight and 5am in a massive evacuation session, and in a public toilet of say 20 bowls shared between 400 people (TCR 20:1). This user NEVER causes another persons usage of the public toilet to be blocked - yet by NTL's definition they are an toilet abuser. This then is the most basic (but not only) flaw with the NTL cap as it currently exists.
I hope that this 'toilet' approach to some issues surrounding contention makes the issues both easier to understand and a little less 'dry'