Revision History
TekSavvy.com - Shared Internet Resources Policy
Show other documents from TekSavvy.com
Version Comments
The "When will the SIR Policy be in effect?" date has been pushed to "early 2017".
The section "4. When will the SIR Policy be in effect?" has been updated.
The following section has been updated: "7. I'm not really feeling the traffic management piece of your SIR Policy. Can I opt out?"
The answer to the question "Is there any other traffic management that TekSavvy end-users should know about?" has been updated concerning blocked ports and SkyFi.
The implementation of the SIR Policy on the Rogers and Videotron platforms has been delayed yet again to "later in October or November 2016" (previously June or July).
The year mentioned under section 8, "I thought TekSavvy was all about net neutrality" has been changed.
The SIR Policy will now be implemented on the Rogers and Videotron platforms later in June or July (previously March or April).
The SIR Policy is to be implemented on the Rogers and Videotron platforms later in March or April (previously February).
The section "When will the SIR Policy be in effect?" has been updated.
This document has been renamed to "Shared Internet Resources Policy" and completely revamped.
Yet again, a new major version has been published.
Another major version has been published.
A new major version has been published.
This is the initial version that ParanoidPaul captured. It is not necessarily the first version of the document.
Shared Internet Resources Policy    
Last updated: 11/03/201611/28/2016

	SHARED INTERNET RESOURCES POLICY ("SIR POLICY") 

	1. WHAT DO YOU MEAN BY "SHARED INTERNET RESOURCES"? 

	TekSavvy provides Internet access. Providing Internet access means making use of shared network resources: the Internet is all about aggregating traffic on shared links, and processing it through shared routing resources.  

	One of the things that makes the Internet awesome is that it puts very few restrictions on what end-users do with their Internet access. That openness, which some people call the "end-to-end principle", gives everyone the freedom to innovate and interact without having to ask permission. It also gives them full access to use as much or as little as they need of those shared Internet resources. Someone who needs more of the shared upstream link will open more sessions in order to demand more of it. Someone who needs less will open fewer sessions and demand less. The Internet is, quite literally, based on sharing. 

	2. WHY DO YOU NEED A POLICY ON SHARED INTERNET RESOURCES? 

	We have always had to decide how to manage our shared Internet resources. Until recently, though, we didn't have a formal policy on how to do so. We did employ a range of measures, like time-of-day incentives and a focus on domestic interconnection. But the centerpiece of our unwritten policy was a very simple concept: provision all the upstream bandwidth we need to minimize congestion. 

	There were two problems with that policy. 

	First, it's not sustainable. No matter how much bandwidth we provision, unusual network events like a new Netflix release or a major new download can and, in many cases, will congest it. Put simply, Internet traffic is growing very, very quickly; peak usage events are shooting past average levels by greater and greater degrees; and every user always has the ability to command extra resources by opening multiple parallel sessions. All of that has the potential to degrade end-users' experiences significantly and increasingly, especially when those experiences involve real-time applications like voice-over-IP or streaming media.  

	Second, even were it possible to provision enough upstream bandwidth that no network event and no user behaviour could saturate it, it wouldn't be sustainable, because that approach is expensive in a way that wastes resources. "Throw more bandwidth at it" is an easy policy to implement, but it forces every user to chip in for all of the bandwidth needed to handle the activity of any user.  

	That is especially hard to sustain when bandwidth is extremely costly. For some years now, TekSavvy has argued that the mandated bandwidth rates we pay to wholesale network access providers have been set far too high. Those wholesale rates are divided into two parts. One is the access part. The other is, in most cases, the size of our point of interconnection with the wholesale carrier, which determines how much capacity gets to the access part - and varies from $10.36 monthly per Mbps (Bell Canada), to $11.81 (Cogeco), to $14 (Rogers), to $20.31 (Videotron). 

	The CRTC has been working steadily to review both access rates and bandwidth rates at the point of interconnection. But it is a slow process. In the meantime, the cost of bandwidth remains high, usage continues to go up, and simply throwing more bandwidth at our shared resources isn't sustainable without hiking prices significantly. 

	The SIR Policy is TekSavvy's attempt to get smarter at how we do things when the network gets congested during the download day (8 a.m. to 2 a.m.) period. 

	3. OKAY, SO WHO IS AFFECTED WHEN BY THE SIR POLICY? 

	TekSavvy is committed to making sure we always provision enough shared bandwidth to serve end-users under ordinary circumstances. That remains the centerpiece of our SIR Policy. Now, though, the SIR Policy also adds two layers of best-efforts traffic management to congested periods that occur during the 8 a.m. to 2 a.m. period, in order to make sure we keep user experience great and prices low. 

	First, we will take steps to make sure that on the rare occasions where there is congestion, it least affects the most real-time protocols, like voice-over-IP and streaming media. Instead, less-real-time applications, like peer-to-peer file transfers, will be more likely to be affected. From what we have seen, even less-real-time applications will not be much affected. Few users--if any--should see any perceptible change in their experience, and more users will see a positive improvement in the use of real-time applications in congested circumstances.  A little bit of management can go a very long way. 

	Second, even when congestion does occur, slightly affecting less-real-time applications, it is those who have used the most bandwidth that day whose use of those applications will be the first to be affected by these measures.  

	We will continue to do our best to make sure that there is no congestion at all. Most of the time, we have been successful. But we are now also acknowledging that, no matter what we do, unusual events or behaviour may cause congestion - and we are working to ensure that, in those circumstances, user experience and user costs are affected as little as possible. 

	4. WHEN WILL THE SIR POLICY BE IN EFFECT? 

	We're phasing it in gradually. In the first phase, the SIR Policy will apply to service provisioned over our highest-cost-bandwidth platforms: Rogers and Videotron. We intended to begin this on December 1, 2015, and provided advance notice to all end-users that would be affected by the first phase of the SIR Policy. However, the SIR Policy is not yet implemented. We do not expect it to be implemented prior to late November 2016,early 2017, on the Rogers and Videotron access platforms. 

	In the following phase, the SIR Policy will apply to service provisioned over our other platforms: Cogeco, Shaw, Bell, and TELUS. We will notify affected end-users well in advance of implementing the policy to these platforms, so that you always know how you are affected. 

	In the same way, as more detailed implementation is logged, we will document it in greater depth. 

	5. WHEN THE SIR POLICY LEADS TO TRAFFIC MANAGEMENT, WHAT KIND OF TRAFFIC DOES IT AFFECT? 

	Traffic management will only take place during the ordinary download day of 8 a.m. to 2 a.m. and, within that period, only when a network is congested. In those circumstances, first to be affected will be the use of less-real-time protocols, by those who have made the greatest use of shared network resources that day. We do not expect anyone to be significantly affected: but that is the sequence in which it will occur. 

	6. HOW DOES THE SIR POLICY AFFECT TEKSAVVY END-USERS' INTERNET EXPERIENCES? 

	We expect that it will improve most end-users' experiences, allow us to stabilize pricing, and lower the cost of most unlimited packages.  

	The SIR Policy calls for traffic management only when there is congestion on a network during the 8 a.m. to 2 a.m. period. Even then, the traffic management affects only those who have made the heaviest use of the network over the course of that day. And even then, the traffic management seen by those users is further restricted to the least-real-time traffic that is least sensitive to millisecond timing. None of these measures involves any rate-limiting or speed caps.  

	For most end-users, those changes will mean better access to shared Internet resources for real-time applications like voice-over-IP and streaming media. A very small subset of our user base will ever experience traffic management - and many of them are not likely to notice it. 

	7. I'M NOT REALLY FEELING THE TRAFFIC MANAGEMENT PIECE OF YOUR SIR POLICY. CAN I OPT OUT? 

Please know that our traffic management measures are as non-invasive and non-intrusive as we can possibly make them. However, users with capped packages, and who'd otherwise be subject to emergency traffic management if peak-period congestion happens, can opt out in exchange for a $5 monthly fee. At present, we only contemplate turning on this kind of congestion management on our highest-cost-bandwidth access platforms (Rogers and Videotron), so it only makes sense to opt out if your Internet is provisioned on one of those access platforms. If your service is provisioned over the Rogers or Videotron access platform, you can opt out by calling us at 1-877-779-1575 or through My Account (myaccount.teksavvy.com). We'll roll out the opt-out functionality across other access platforms, at the same pace as we roll out any required traffic management itself. 

	 Now, recall that one of the reasons we have implemented this formal SIR Policy is to get smarter about how we control shared costs, in order to be fair to our whole user base: hence the charge. For that reason, users with unlimited packages will not have the ability to opt out of traffic management.   

	8. BUT THEN WHAT HAPPENS TO ZAP THE CAP? 

	We are phasing out the Zap the Cap option that has allowed many of our users to turn their capped plans into unlimited plans in exchange for accepting slower-at-peak speeds. Instead, we are dropping the price of some of our unlimited plans significantly. A smarter SIR Policy lets us do that. 

	9. I THOUGHT TEKSAVVY WAS ALL ABOUT NET NEUTRALITY. 

	We are. Back in 2008, TekSavvy organized a rally on Parliament Hill in support of network neutrality. Our views haven't changed. We're simply an intermediary, and we have no business telling our users what to do or snooping on what you actually do. We haul bits to and from end-users, in as neutral a fashion as possible. 

	But that doesn't mean we want to charge you higher prices than we need to, nor cause users of real-time applications to experience congestion when they don't have to. Our SIR Policy formalizes our approach to optimizing those things. TekSavvy's measures are applied only where necessary, and look only at classes of traffic. We don't look at content, don't look at where traffic is coming from, and don't look at where it's headed. 

	10. WHAT ARE THE PRIVACY IMPLICATIONS--DOES THIS MEAN TEKSAVVY IS TRACKING, RETAINING, OR REPORTING WHAT I DO ON THE INTERNET? 

	At TekSavvy, we take privacy very seriously. We have no interest in tracking what you do. We have no interest in looking at any of the content that passes over your network. We certainly have no interest in monitoring your online activities. While we respect all of our legal obligations, we're an intermediary. That's all. 

	We do measure your traffic volume, because we make capped plans available as well as unlimited ones. We do analyze protocols at the aggregate level, because we want to make sure congestion doesn't cripple real-time applications, and that we don't waste money buying bandwidth we don't need. But all of that analysis is done on the fly, and discarded immediately. It's none of our business. 

	11. IS THERE ANY OTHER TRAFFIC MANAGEMENT THAT TEKSAVVY END-USERS SHOULD KNOW ABOUT? 

	Yes.  

	First, we block traffic associated with three ports, 25 (outbound), 53 (inbound), and 1900 (inbound), which are typically associated with network attacks. Outbound port 25 traffic, associated with TCP SMTP communications, is vulnerable to concurrent connection attacks and spam abuse, so only connections to TekSavvy's SMTP server are permitted. Inbound port 53 traffic, associated with UDP DNS server communications, is vulnerable to DNS denial-of-service and related attacks. Inbound port 1900 traffic, associated with Microsoft's Simple Service Discovery Protocol, is similarly vulnerable to a range of distributed DOS-style attacks. If you have a specific requirement requiring that these ports be maintained open, please contact us.  

	Second, most of our end-users are provisioned over wholesale network access acquired from third-party carriers. And most third-party carriers have chosen to deliver that access at the Internet layer, which effectively makes our end-users subject to the same Internet Traffic Management Practices that those incumbents impose on their own retail end-users, as follows.  

	 Bell has indicated that its wholesale network access is not currently subject to any traffic management policies. 

	  Cogeco indicates that its wholesale network access is subject to deep packet inspection with the single purpose of managing peer-to-peer applications. Content analysis is restricted to traffic classification only for traffic management purposes on upstream traffic. No content records are maintained. No Internet traffic management measures are applied in the downstream direction. 

	  Rogers indicates that its retail network access is subject to what Rogers describes as "a variety of network management techniques. These techniques have evolved as the Internet has changed" to limit spam, viruses and other security threats. We believe that the same network management techniques are applied to Rogers' wholesale network access.  

	  Shaw indicates that its network access is subject to Internet traffic management practices that reduce, to 80 Kbps per end-user, upstream bandwidth allotted to "P2P applications completing non-real-time file transfer activity" within serving area nodes experiencing "upstream congestion". These management practices are "not applied to downstream data transfer, real-time interactive or time-sensitive Internet applications." 

	  TELUS has indicated that its retail network access is subject to Internet traffic management in some areas where demand exceeds available network capacity. These include areas where TekSavvy has end-users, and we believe the same traffic management is applied to TELUS's wholesale network access. When they identify congestion in these areas, TELUS will redistribute network capacity across all users by temporarily reducing the speed of the few heavy users. This is intended to provide better speeds to the majority of other customers.

	 Videotron has indicated that its wholesale network access is, at speeds of 120 Mbps speeds or higher, subject to upload measures that Videotron has explained are to prevent users from congesting shared "upload channels" serving a few dozen modems. Every 15 minutes, a system checks the usage rate for each upload channel. If the usage rate has reached a threshold beyond which congestion is imminent, the system identifies the 120 Mbps and 200 Mbps modems on that channel that have uploaded a statistically significant amount of data. Uploading from these modems is then momentarily given lower priority. Depending on the severity and duration of the congestion, uploading speed may be slowed for these modems. When the amount of data uploaded by the modem is diminished or the transmission of data goes back to a rate that does not cause congestion, uploading will return to its usual priority. 

	End-users whose Internet service is provisioned over TekSavvy's own SkyFi access, a fixed-wireless access service whose users share limited spectrum, are subject to inbound traffic blocks on ports 21 (FTP), 22 (SSH), 53 (DNS), 80 (HTTP), and 443 (HTTPS). This prevents FTP, SSH, DNS, and Web servers from being operated from behind a consumer fixed-wireless connection, maintaining overall service quality within the shared spectrum.

If you'd like to be notified when TekSavvy.com makes updates to documents like this, choose which ones you'd like to subscribe to today (it's free!).