Summary: CRTC PN 2008-19; ISP Traffic Managment in Canada

/ Christopher Parsons

As someone who is academically invested in how the ‘net is being regulated in Canada, I’ve been following the recent CRTC investigation into Internet management practices and regulation with considerable interest. Given that few people are likely to dig though the hundreds of pages that were in the first filing, I’ve summarized the responses from ISPs (save for Videotron’s submissions; I don’t read French) to a more manageable 50 pages. Enjoy!

Update: Thanks to Eric Samson and Daniel for translating Videotron’s filings – you guys rock!

13 thoughts on “Summary: CRTC PN 2008-19; ISP Traffic Managment in Canada”

  2. It’s ‘Videotron’, not ‘Videotronic’, and I can take a look at the papers and give you the gist of it if you want. Let me know. I’m already reading a bit of them, what with being one of their customers…

    Like

  3. I’m not sleeping – might as well. I’ll go document by document.

    Basically, Videotron is saying:

    Cover letter – that traffic, bandwidth, peak usage and methods used to alleviate usage at peak times are confidential matters.
    1 – Traffic is confidential, percentage of traffic per protocol is not available, peak periods have not changed but are defined in confidence, and downloading will evolve faster than uploading in 2009-2011, mostly because of video content, and the fact that users can better control their upstream usage. Percentages are confidential.
    2 – Everything is in confidence, except question D, which is answered as “information unavailable”.
    3 – Congestion is defined as (REDACTED), but policies do not vary whether this is a end-user or reseller of Videotron’s services.
    4 – Criteria used to define congestion have not changed between 2006-2008.
    5 – Videotron’s hybrid cable and fiber network is composed of three elements: Internet links to the outside world, a federating network, and DOCSIS canals serving one or more cells, each cell corresponding to a certain quantity of users. All these ressources are shared.
    6 – When congested:
    * Outbound Internet – Videotron adds capacity (REDACTED)
    * Federating network – Videotron adds capacity (REDACTED)
    * Upstream DOCSIS canal – Videotron considers it congested when (REDACTED)
    * Downstream DOCSIS canal – (REDACTED)
    7 – Traffic shaping practices have not been changed in 2006-2008, but specific solutions used to eliminate congestion have evolved. These are (REDACTED).
    8 – “Videotron does not use Internet traffic managing technologies like deep packet inspection.” That’s it.
    9 – Basically retells once and again that Videotron does not throttle Internet connections, but imposes up- and download quotas to its users, thereby creating financial incentives for users to watch their Internet usage. It is the only thing Videotron does in order to stop a minority of users from taking up a majority of the bandwidth. Even after the user reaches his monthly transfer limit, speed does not change.
    10 – Videotron has no plans to change this policy.
    11 – Videotron does not use traffic managing technologies because “For the moment, Videotron is satisfied by the results obtained by management practices based on transfer limits and overusage surcharges.”
    12 – A dramatic surge in peak-time usage by all or some of Videotron’s customers might make Videotron rethink its policy on this matter.
    13 – “No.” (There are no other techniques used by Videotron to manage Internet traffic.)
    14 – See numbers 11 and 12. We can not speculate on any other reasons that would make us rethink our policy on this matter.
    15 – Videotron informs its customers of the monthly limits of their subscription. All resellers get the same information, and it is up to them to communicate them, or not, to the customer.

    De nada.

    Like

  4. I went through the Videotron documents. For all questions that require data about traffic and consumption, the reports were filed in confidence. Also, Videotron did not modify the way it regulates traffic since 2006.

    To me, the most important questions were the following :

    Q5-Q6 : Internet links : Videotron adds capacity.
    Federative network : Videotron adds capacity.
    Downstream DOCSIS channels : Filed in confidence.
    Upstream DOCSIS channels : Filed in confidence.

    Q8 : Videotron states that they do not use any traffic management technologies such as DPI.

    Q9 : Videotron imposes data transfer limits to it’s clients. They do not reduce the speeds once the limit is reached, they simply charge an extra fee to the customers. This is an incentive to encourage end-users to keep their consumption reasonable. The limits are openly posted on the companie’s website : http://www.videotron.ca

    Please feel free to correct me on any point as all the previous information was translated from French documentation.

    For those who don’t know about it, Videotron is a major ISP in the province of Quebec. They use a coaxial/fiber cable network and, I don’t have any numbers in front of me, but they might be bigger than Bell in metropolitan areas.

    Like

  5. First of all – thank you.

    Did you come across any indication as to what sort of peak capacity per user the ISPs are building in Canada?

    As an avid user I suggested in my submission such as it was that the relevant Trade Descriptions Act ought to be enough to force ISP to publish their planning rules as a guide to how end to end services ought to work!

    The lack of transparency does not look clever overall.

    Like

    • @ Mike: Off the top of my head, I don’t believe that the filings stated this information, at least not to the public. Details may have been filed in confidence. Only TELUS gave (more or less) unfiltered access to the capacity used by the 5% and 10% categories in GB, but didn’t note a max capacity that a user can claim before being removed from the service.

      Like

Comments are closed.