 MattAll noise, no signal.Premium
join:2003-07-20
Jamestown, NC
kudos:12
1 edit | reply to fAcEtIOUs
Re: TekSavvy wants no throttling of anyone said by fAcEtIOUs:said by Matt:The congestion has actually gone UP since they implemented throttling. Has it? Who says? If you look at the numbers they submitted, their measure of "packet loss" has gone from a steady 2-3% before DPI, to a steady 8% in the months since they implemented DPI.
The numbers are there in there filing. Everyone on the internet has picked them apart and called BS. Their DPI equipment is causing MORE issues, which is to be expected due to the nature of DPI.
»arstechnica.com/news.ars/post/20···ent.html
Quote:
But the congestion rates weren't high before throttling began. One year ago, in May 2007, the aggregation link was congested 1.8 percent of the time, the BAS was congested 0.4 percent of the time, and the backbone experienced congestion 3 percent of the time.
Data source: Bell Canada
At the DSLAM, though, congestion hovered around 4 percent last year. From October 2007 to May 2008, when Bell was rolling out its traffic-shaping gear, congestion at the DSLAM picked up dramatically, and Bell set three new congestion records (going from 7.8 percent to 8 percent to 8.2 percent). |
|
|
|
| said by Matt:If you look at the numbers they submitted, their measure of "packet loss" has gone from a steady 2-3% before DPI, to a steady 8% in the months since they implemented DPI. The numbers are there in there filing. Everyone on the internet has picked them apart and called BS. Their DPI equipment is causing MORE issues, which is to be expected due to the nature of DPI. Your conclusion is based on a logical fallacy(If B follows A, therefore A caused B). Because congestion followed the installation of DPI equipment you can't conclude that it caused congestion. There are other variables involved including rapidly growing bandwidth demand throughout the period(as shown in the submitted supplement(  080623_CAIP ···R(2).zip 352538 bytes
). Without DPI equipment, congestion could very well have been even worse.
So the data doesn't support your conclusion. It doesn't exclude it either.
Also, the data collected by Bell was too coarse to come to any guaranteed conclusions. See my post on that issue: »Low % congestion #'s could still indicate problem though
In conclusion: the data submitted could have been more comprehensive. But I'll take the opinions of the network engineers actually running the network rather than opinions of those who are basing their analysis on only the small subset of data submitted to the CRTC.
--
My BLOG .. .. Internet News .. .. My Web Page
Ask yourself one question: 'Do I feel lucky?' Well, do ya punk? |
|
MattAll noise, no signal.Premium
join:2003-07-20
Jamestown, NC
kudos:12 | said by fAcEtIOUs:In conclusion: the data submitted could have been more comprehensive. But I'll take the opinions of the network engineers actually running the network rather than opinions of those who are basing their analysis on only the small subset of data submitted to the CRTC. And I think that is where you and I differ on opinion. I think that Bell knows they have congestion, knows what it will cost to fix that congestion, but they also wants to eliminate their CAIP competitors, while reducing congestion for their forthcoming video service.
From a business perspective, DPI gear solves a LOT of problems for Bell and is cheaper than performing the required upgrades across their service area. However, if they admitted why they really rolled it out, they'd be admitting to breaking Canadian telecommunication laws.
I think in this case there are too many "coincidences" pointing to ulterior motives for there not to be any and Bell's attitude of "We say it, so that's how it is." doesn't help their cause. |
|
 mazhurgPremium
join:2004-05-02
Portage La Prairie, MB
 ·MTS
| reply to fAcEtIOUs
said by fAcEtIOUs:In conclusion: the data submitted could have been more comprehensive. But I'll take the opinions of the network engineers actually running the network rather than opinions of those who are basing their analysis on only the small subset of data submitted to the CRTC. What we have heard so far is not from the network engineers but from the spinsters. The numbers submitted (in the public documents) are, as you say, too vague. However, it stands for reason that if congestion was really the case, don't you think Bell would have trumped the real numbers loud and clear? That they did not, and that they are using dubious analysis (and results) to try to prove congestion indicate that there probably is not. |
|
Reviews:
 ·RoadRunner Cable
1 edit | reply to fAcEtIOUs
said by fAcEtIOUs:said by Matt:If you look at the numbers they submitted, their measure of "packet loss" has gone from a steady 2-3% before DPI, to a steady 8% in the months since they implemented DPI. The numbers are there in there filing. Everyone on the internet has picked them apart and called BS. Their DPI equipment is causing MORE issues, which is to be expected due to the nature of DPI. Your conclusion is based on a logical fallacy(If B follows A, therefore A caused B). Because congestion followed the installation of DPI equipment you can't conclude that it caused congestion. There are other variables involved including rapidly growing bandwidth demand throughout the period(as shown in the submitted supplement( [att=1] ). Without DPI equipment, congestion could very well have been even worse. So the data doesn't support your conclusion. It doesn't exclude it either. Also, the data collected by Bell was too coarse to come to any guaranteed conclusions. See my post on that issue: » Low % congestion #'s could still indicate problem thoughIn conclusion: the data submitted could have been more comprehensive. But I'll take the opinions of the network engineers actually running the network rather than opinions of those who are basing their analysis on only the small subset of data submitted to the CRTC. This is now way shape or form proves any of your argument. Your strategy of "out posting" will not dig you out of this hole. Please provide evidence of Bell Canada's congestion. All of Canada is eagerly awaiting your proof TK. |
|
| reply to mazhurg
Exactly,
If congestion were the real problem they would have 2 filling cabinets full of memo's, graphs, and other internal documents to back up these claims. One chart produced in Microsoft word reaks of BS. TK, in all your years as a network designer did you make business cases based on one chart produced by someone with a 10th grade education? |
|
 root9
join:2005-04-08
Kitchener, ON | reply to fAcEtIOUs
Hahaha , you've been busted TK, by general public yet.
You don't know since you don't have access to Bell's network, settings, DPI tech. and internal politics.
You don't know because you don't have all the information.
You can assume all you want and you know [as with your experience] Bell's provided evidence is obviously flawed. As a matter of fact it's outright lies in many areas.
To boot it's illegal in the worst ways.
I doesn't seem like you know much about Canadian laws, do you?
--
Please engage eyeballs and retain functional brain before operating fingers. |
|
