how-to block ads
|
|
Share Topic
 |
 |
|
|
|
 matteiModerated, now muzzled
join:2001-03-19
Canada | reply to Krispy
Re: cogeco's new monitor on the fritz? For a time-based scenario (polling, etc):
Cogeco Lite Residential Service:
- 1 GB-per-overage charge = 8,388,608 Kb-per-overage charge
- 8,388,608 Kb-per-overage charge / "Up to" 640 Kbps = "At least" 13,107 seconds/overage charge
- "At least" 13,107 seconds/overage charge = A maximum of 218.5 minutes between updates for best effort bill accounting
Cogeco Lite Plus Residential Service:
- 1 GB-per-overage charge = 8,192 Mb-per-overage charge
- 8,192 Mb-per-overage charge / "Up to" 3 Mbps = "At least" 2731 seconds/overage charge
- "At least" 2731 seconds/overage charge = A maximum of 45.5 minutes between updates for best effort bill accounting
Cogeco Standard Residential Service:
- 8,192 Mb-per-overage charge / "Up to" 10 Mbps = "At least" 819.2 seconds/overage charge
- "At least" 819.2 seconds/overage charge = A maximum of 13.7 minutes between updates for best effort bill accounting
Cogeco Pro Residential Service:
- 8,192 Mb-per-overage charge / "Up to" 16 Mbps = "At least" 512 seconds/overage charge
- "At least" 512 seconds/overage charge = A maximum of 8.5 minutes between updates for best effort bill accounting
Otherwise, the threshold for pushing an update should be every billing unit change. | |
| You're using numbers that assume a person is using their connection all out all the time. You also forgot upload. | |
matteiModerated, now muzzled
join:2001-03-19
Canada | said by InfoRanger :
You also forgot upload. You're right.
I used the large download numbers to estimate sampling times. That's only one side of the customer's asymmetric traffic passed (last mile) on Cogeco's HFC network. The larger side. I was being lazy when responding to comments about the Bandwidth Monitor update interval currently being much larger than numbers I feel are appropriate. Including upstream would decrease the intervals I presented.
I'm also guilty of not mentioning upstream in the other numbers post I made in this thread. Inclusion would decrease those "At least" figures as well.
The "all out" numbers are used exactly because they're maximums. The perspective I'm using for traffic accounting, a customer's, is everything passing through the modem/access device adds to your total (historically true, I think). That it might not be possible to saturate your connection profile due to network conditions at a given time is irrelevant. It must be possible some of the time - Cogeco does sell a service using these numbers. In fact, from most accounts, they provide a level of service close to or meeting their advertising. As for individual usage patterns, they're just that - individual. You can provide satisfactory accounting with a minimum sampling frequency obtained using the approach I've taken, use a similarly/more accurate alternative or deal with increased bill disputes and churn. Maybe that's the point?
I have no idea how Cogeco's customer facing Bandwidth Monitor system is implemented but was surprised to read an infrequent update complaint/request taken as a suggestion and not a problem report. By drawing attention, I was hoping for a discussion or a "don't worry, it's current" response that I could verify some time. | |
matteiModerated, now muzzled
join:2001-03-19
Canada | I thought I'd add...I did a bit of reading today. Intent, direction, and concerns seem clear enough:
Usage Billing ApproachThe usage billing approach was and is still not implemented for the following reasons: - Usage billing requires extensive modifications to Cogeco’s IT systems. At that time, these modifications were considered too costly and too long to implement with respect to the benefit and the urgency of the situation. Cogeco is however currently in the process of implementing these required IT system modifications and plans to start billing its customers for IP usage above the bandwidth limitations mentioned in Cogeco’s HSIAUP later this year. See response to interrogatory Cogeco(CRTC)4Dec08-10;
- In 2001, the DOCSIS infrastructure was not built to perform usage billing and the cable modem counters were not deemed reliable enough to use as a metric for billing purposes. The development of the Internet Protocol Detail Record (IPDR) was in its infancy. In fact, IPDR was included in the DOCSIS Certification Waves only in May 2005. Lately, CMTS vendors have implemented IPDR in their equipment and IPDR Billing systems are now available on the market. See the IPDR milestone table available on the TM Forum at:
»www.ipdr.org/BestPracticesStanda···ome.html
In any event, the implementation of traffic management equipment enabling Cogeco to monitor and manage its network was necessary. Usage billing was not and still is not considered as a sufficient solution to enable Cogeco to meet its commitment to provide each user with prompt access and fast connection to the World Wide Web, due primarily to the fact that the length time from action to bill would allow subscribers to “abuse” the network for a significant time period without immediate redress for adjacent users. To this end, Cogeco should have the ability to monitor usage and manage bandwidth in a timely fashion in order to protect its network and ensure a fair usage among its customers as well as compliance with applicable terms of use. An excerpt from
2009-01-13 - Cogeco Cable Inc.
Description: Pursuant to paragraph 17 of Telecom Public Notice CRTC 2008-19, Review of the Internet traffic management practices of Internet service providers, Cogeco Cable Inc. (Cogeco) submits on behalf of Cogeco Cable Canada Inc. its responses to the Commission interrogatories.
»www.crtc.gc.ca/public/partvii/20···5754.zip | |
|
