CAIP on Bell Canada traffic throttling: final

p2pnet news view Freedom | P2P:- The CBC has picked up on the latest, and final, CAIP submission to the CRTC over the Bell Canada traffing throttling scandal.
Bell implies P2P is the, “motive force behind a wave of ‘bandwidth hungry’ applications that will soon overwhelm global networks in the absence of aggressive measures taken against it,” says the submission.
“It also states that all P2P data transfers should be considered the postal equivalent of ‘bulk’ or ‘lower urgency’ mail, and that such transfers are never ‘time-sensitive’ to the same extent as web browsing and other applications.”
p2pnet posted the full document late yesterday and today, “The Canadian Association of Internet Providers, a group of 55 companies that rent portions of Bell’s network to provide their own broadband services, made its last plea Wednesday to regulators to force Bell to end its speed throttling,” says the CBC.
In November Bell started using deep packet inspection (DPI) technology to identify what its customers were using their internet connections for, then started slowing peer-to-peer (P2P) applications, it says, continuing:
“The company extended the practice to CAIP members in March, which prompted the group’s complaint to the Canadian Radio-television and Telecommunications Commission in April.
“In its submission on Wednesday, CAIP said Bell’s defence for throttling - that the company is only slowing P2P downloads, which still get to the user “eventually” - is discriminatory and anti-competitive.”
Ontario-based ISP TekSavvy was among the first, if not the first, to protest Bell’s attempt to shackle its users and impose a block on net neutrality.
“If you believe in what CAIP believes in now’s the time to educate the world,” says CEO Rocky Gaudrault (right) on dslrports, adding »»»
The old way of doing business, where ethics and profits are at odds, must cease to exist! Being fair, truthful and transparent is the only way to go going forward!
Tell your friends and family so that this case be used as an example of a serious need to rewrite how things are run.
We, the clients, in the end have the power and if the decision-makers and politicians, elected by these same clients, are told unanimously that corporate bullying must stop, then backdoor lobbying and spin tactics won’t matter anymore.
Change will come ……..
The CBC has the CAIP document stating, “Postal service customers have the freedom to decide for themselves the urgency of their packages, and to pay the postal service a fee based on how quickly they want their packages delivered.
“Bell’s high-handed imposition of traffic management is more appropriately analogized to a postal service that opens each package, decides according to its own priorities how important the contents are, and delivers it at a speed of its own choosing, notwithstanding the needs or intentions of package senders and recipients.”
A spokesperson for Bell didn’t return a request for comment, but the company has said it only uses DPI to determine the type of traffic, and does not probe any deeper,” says the CBC.
“The CAIP submission was the final step in a public CRTC investigation, which began in May, into whether Bell has violated the Telecommunications Act by illegally changing its terms of wholesale service,” it says. “The commission is expected to rule on the case in September.”
CAIP says ……..
In its submission, “Bell’s initial stated rationale for deploying traffic shaping was to rein in the bandwidth usage of an ill-starred five per cent of internet users who, according to Bell, were ruining it for everyone else,” continuing »»»
These users, Bell claimed, employ P2P file transfer applications in a manner that consumes a disproportionate amount of bandwidth, creating negative impacts on other users due to congestion on the network.
Contrary to Bell’s assertions, the use of P2P applications to transfer content through the Internet does not consume more bandwidth than the use of more traditional client-server architectures. Bell’s assertions in this regard, which echo the rhetoric of DPI equipment manufacturers and TELUS and Rogers, do not fairly represent P2P technology and appear calculated to mislead both the public and the Commission.
At paragraph 40 and following of Bell’s Final Answer, Bell purports to explain how a P2P user manages to consume a disproportionate amount of bandwidth. Bell, citing Rogers’ colourful language, states that P2P is “designed to overwhelm other traffic.” Bell goes on to state that a P2P application will open up to 40 to 100 TCP sessions in an effort to transfer data as fast as possible using multiple sources. Bell then reproduces in Figure 4 a diagram of a “pipe” depicting how P2P user A “hogs” a disproportionate amount of bandwidth as compared to non-P2P-user B.22
Bell fails to mention a number of very important assumptions or facts regarding Figure 4 and how TCP works:
(a)The reality in relation to the end-user of a GAS customer is that the total amount of bandwidth used by the end-user is limited to the GAS subscription bandwidth (e.g., 5 Mbps downstream/800 kbps upstream), regardless of how many sessions the user has open. Thus, the end-user has a real cap on the amount of bandwidth that he or she can upload or download.
Bell’s Figure 4 assumes that there is no bandwidth limit at the source or destination of the transmissions.
(b)In reality, P2P downloads are made from other end-users. In such case, bandwidth is limited at both the source and destination end. On Bell’s network, each of these uploading connections for end-users of GAS customers is limited to a maximum of 800kbps. The download connection is also limited by the GAS tariff. Thus where the download connection is for the 5 Mbps GAS service, at the very most, the downloader could sustain, assuming that the uploading connection is operating a full speed, is six connections or sessions.
Bell’s Figure 4 implies that a P2P user may download or upload using a limitless number of sessions.
(c)Bell fails to disclose that in the original diagram prepared by George Ou upon which Figure 4 is based, the large green cutaway pipe represent a congested network link.23
(d)Bell fails to disclose that under no circumstances, congested or otherwise can a P2P application take up more bandwidth than that which is defined by the speed at the source and destination of the communication.
(e)George Ou concedes that for purposes of that diagram, “Downloads don’t matter. Downloads aren’t congested.”24
When there is congestion in the network, the TCP protocol alleviates the congestion by instructing communicating computers to reduce the amount of data entering the network. This is done amongst all connections flowing through the congested portion of the network.
While P2P applications that are being uploaded may take up more bandwidth than non-P2P connections when the connection is congested (because the P2P application does not self-constrain itself to the same extent as other TCP-based applications), as stated above, even in this circumstance, the P2P application cannot exceed the bandwidth allocated at the source and destination of the communication by the ISP or in our case the GAS tariff. When there is no congestion, the phenomenon described by George Ou in his diagrams will not be observed. Only once a pipe hits the point of congestion will TCP “stacks” drop packets and force non-P2P connections to reduce the speed of transmission, thus alleviating congestion.
With the foregoing clarifications, it should be evident that P2P is not the threat that Bell makes it out to be. Furthermore any bandwidth effects associated with P2P are entirely contingent on whether or not congestion actually exists over a given link. CAIP devotes Part V of this Reply to de-bunking that congestion myth. If the link is not congested then none of Bell’s claims regarding P2P’s disproportionate use of bandwidth apply. It is necessary to reiterate that through its GAS service, Bell is merely providing competitors with last mile connectivity – an “opaque pipe” between the competitor’s point of presence and the household of an end-user. With P2P, a file is downloaded in segments simultaneously from a number of sources. Although a number of transfers may be occurring at the same time, the total volume of traffic going over the last mile is the same whether the file is loaded from or to a source or a number of sources. As such, in terms of traffic, it makes no difference whether content is transferred through P2P or any other protocol. It is misleading to argue that a movie or television program takes up more bandwidth over the last mile because an individual chooses to access that content through P2P.25
There is therefore no justification for throttling the content of GAS customers simply because some of this content is distributed through P2P applications.
Furthermore, while global bandwidth usage is growing exponentially, a recent study by Cisco (cited at p.20 of Bell’s Final Answer) predicts that the relative proportion of P2P traffic on the Internet will actually decrease over the next four years.26 This raises the question – which Bell does not satisfactorily answer – as to why Bell chose to discriminate against P2P in its traffic shaping program. CAIP notes that P2P, because of its use by some persons to share copyrighted content, may present a politically-easier target for Bell than other kinds of data.
CAIP also points out that in trying to justify its use of DPI (deep packet inspection), Bell attempted to characterize P2P file sharing as an “unfair” or even “rogue mode” of transferring data across networks.
“In particular, Bell implies that P2P is the motive force behind a wave of ‘bandwidth hungry’ applications that will soon overwhelm global networks in the absence of aggressive measures taken against it,” says the submission.
“It also states that all P2P data transfers should be considered the postal equivalent of ‘bulk’ or ‘lower urgency’ mail, and that such transfers are never ‘time-sensitive’ to the same extent as web browsing and other applications.”
It goes on »»»
In response to these misleading representations, CAIP notes that from its early roots as a marginal and even counter-culture method of data distribution, P2P applications, especially BitTorrent, have become fully mainstream, fully commercialized and fully endorsed by a number of major movie studios and content creators. For Bell to characterize BitTorrent as the tool of bandwidth abusers is both short-sighted and manifestly incorrect. For instance, many of the major U.S. media content providers, including Fox, MTV, Paramount and Warner Brothers have endorsed BitTorrent as a method for distributing their offerings.28
The European Union has also recognized the value of P2P technology and has announced that it will invest €14 million in “P2P-Next”, an “open source, efficient, trusted, personalized, user-centric, and participatory television plus media delivery mechanism with social and collaborative connotation using the emerging Peer-to-Peer (P2P) paradigm, which takes into account the existing EU legal framework”.29
A recent survey conducted by the Canadian Internet Policy and Public Interest Clinic (CIPPIC) helps to demonstrate the self-serving nature of Bell’s imposition of meaning on P2P. Hundreds of legal applications and services in every conceivable area of public and private online life utilize P2P transfers.30 Some of these applications are increasingly important to the cross-pollination of ideas in art, music, politics, business and the technological and sociological development of the internet itself. Others, like the massive files transferred by Dr. Andrew D. Fernandes at the Schulich School of Medicine at the University of Western Ontario, are critical to research in other fields.31 A catalogue of many of these mainstream uses appears in an Appendix of CIPPIC’s submission on behalf of the Campaign for Democratic Media.32 Included among these uses are the following:
and(a)BitTorrent – the most popular P2P file sharing protocol is used increasingly for the distribution of many types of content, such as:
(i)films and music by artists such as Nine Inch Nails,
(ii)television programs by the Canadian Broadcasting Corporation (CBC), and
(iii)computer software (e.g. Linux operating system) and video games for use on the Sony Playstation 3;
(b)Television, film and music distribution channels – Numerous distribution channels based on P2P technology have emerged, including: Babelgum, Blinkx BBTV, the British Broadcasting Corporation (BBC)’s iPlayer, GridNetworks, Joost, Livestation, Mashboxx (soon to be released, authorized by major record labels such as Sony BMG Music and EMI Music), Open Media Network, Pando Networks, RedSwoosh, Realtime Rentals, and Tribler Video Streaming,
(c)Private file sharing – P2P allows users to transfer large files in an efficient and secure manner;
(d)Privacy protected communications – Allow users who wish to avoid advertising targeted at behaviours, citizen journalists in countries without safe access to media, law enforcement establishing anonymous tip lines, activists, and whistleblowers to communicate securely via the Internet;
(e)Real-time communications (e.g. popular VoIP application, Skype);
(f)Web search engines, such as YaCy and FAROO;
(g)Platforms for business collaboration (e.g. Collanos Workplace); and
(h)Adobe Flash Player – the next version of Flash Player which will enable users to stream audio and video to other peers.
Bell admits it’s been running DPI equipment on its Sympatico customer base since an “unspecified point in time” last year.
Calling the effort “technical trials,” its says it began to, “apply throttling measures against its Sympatico customers on a commercial basis, reducing the transfer speeds of P2P file transfers and other encrypted traffic to approximately 30 KB/s between the hours of 4:30 p.m. and 2:00 a.m. daily,” also stating »»»
Concerned by this turn of events, Bell’s GAS customers met with Bell on 23 November 2007 and asked Bell directly whether it intended to use DPI on its wholesale ADSL access services. The Bell representatives present at the meeting stated that Bell would only use the technology in conjunction with its retail Sympatico high speed internet service.
The reaction from Bell’s retail Sympatico customers was felt almost immediately. Indeed, the negative feedback from this community of users expressed in the blogosphere has been intense and growing ever since Bell began throttling its own customers.
It is clear from BCE’s annual reports that Bell knew full well that if it would lose customers to competitors if it introduced traffic shaping technologies. In its 2007 Annual Report, Bell indicates that it understood that the discontinuance of such unlimited usage plans using DPI as a precursor would likely cause Bell customers to migrate to competitors.
Given this anticipated response by Bell’s retail customers, and notwithstanding its earlier promise to competitors that it would not shape their GAS traffic, Bell began to shape its competitors’ GAS traffic on 14 March 2008, without any notice of any kind. Competitors, therefore, had no opportunity to test the impact of Bell’s traffic shaping technology on the services that they deliver to their end-user customers. No information was provided to GAS customers as to exactly what changes Bell has made.
Beginning on 14 March 2008, CAIP’s members were therefore subjected to unexpected network disruptions. As discussed below, the evidence in this proceeding shows that their users were affected without notice. Like Bell’s Sympatico customers, its GAS customers did not receive any prior notification from Bell that it intended to introduce traffic shaping measures on their GAS traffic. In fact, not only was no notice given by Bell to these customers, it also failed to provide any details to these customers as to the type of traffic that its DPI boxes would be programmed to detect and throttle. Needless to say, Bell never asked these customers in the period leading up to 14 March 2008 to help develop alternative methods for handling its alleged network congestion issues.
After numerous complaints by independent ISPs, Bell belatedly issued a statement on 28 March 2008 confirming that it had unilaterally decided to expand its traffic shaping practices to its wholesale customers.
Almost a month later, Bell issued a two-page document entitled “DSL Traffic Management - Frequently Asked Questions (FAQ),” which was not even received by all of CAIP’s members who subscribe to Bell’s GAS service. Moreover, the document was sent out more than one month after Bell commenced the throttling of its competitors’ traffic.
From a technical perspective, Bell’s FAQ document is almost entirely bereft of any useful information regarding its traffic shaping equipment and network protocols. Moreover, when asked to provide this type of information, Bell has been extremely evasive and short on technical details. Bell’s technical support staff appear to have a very superficial knowledge of the DPI equipment in Bell’s network, yet they refuse to allow competitors to speak to the equipment vendor directly. In fact, CAIP’s members have reported that Bell will not permit its DPI equipment vendor to troubleshoot or even speak to competitors who have been affected by Bell’s throttling practices.
Even to this day, Bell has provided virtually no information as to what measures/instructions it has put in place in its routers, switches and other network equipment to affect the type of traffic that its DPI boxes have been programmed to filter and shape. Nor has Bell provided any notice of the rollout schedule for its DPI boxes or the locations in which DPI boxes would be deployed. It has been left entirely up to the carriers and other telecom service providers that interconnect with Bell to determine all of these matters.
Competitors require this information in order to properly handle the trouble reports that they receive from their own customers and to figure out, more generally, how best to manage their networks in the face of Bell’s traffic shaping measures.
In its public comments and its submissions to the Commission, Bell has remained unapologetic and non-conciliatory for the high-handed and abrupt manner in which it instituted this major change to its relationship with its wholesale customers. The company has claimed the right to unilaterally and secretively inspect competitors’ GAS traffic as well as to alter a tariffed service mandated by the Commission for specific use by competitors.
CAIP’s submission focused primarily on issues involving regulatory requirements.
The CRTC only has control over the Telecom Act and associated regulations and as Gaudrault says, “Change will come.”
With that in mind, the CRTC has promised to examine net neutrality.
So stay tuned.
Jon Newton - p2pnet
.
.Stumble It!
New York Times - New worm transcodes MP3s to try to infect PCs, July 18, 2008
Subscribe to p2pnet.net | | rss feed: http://p2pnet.net/p2p.rss | | Mobile - http://p2pnet.net/index-wml.php
Net access blocked by government restrictions? Use Psiphon from the Citizen Lab at the University of Toronto. Go here for details. Download here.







July 24th, 2008 at 7:59 am
I get the feeling this ins’t final…..
July 24th, 2008 at 8:00 am
ins’t = isn’t (I failed grammer skool)
July 24th, 2008 at 9:29 am
Spelling school too!
July 24th, 2008 at 9:33 am
I learnded from Jon, not my fault. :p
July 24th, 2008 at 9:57 am
Thenks, gies
Cheers!
July 24th, 2008 at 10:52 pm
hahahaha spiling mistaces maid bi lusers!!!
July 25th, 2008 at 12:40 am
What Bell is doing is the equivalent of overbooking a flight and then making paying passengers wait for a later flight without compensation or apology. The customer is given an upload and download bandwidth as the basis of the monthly price he pays. (S)he ought to be able to use that bandwidth fully without interference.
July 25th, 2008 at 1:40 pm
@JuDGe
yup
July 25th, 2008 at 9:08 pm
attempt #3 at posting something in this thread.
I had posted my comments on the CAIP paper in the other topic. Perhaps we’re not allowed to enter “Hugh Are Elle” in this system and why my previous attempst nevert showed up in this thread.
July 26th, 2008 at 6:06 am
I had a similar experience.
When I posted i got a page that said something like. “slow down you are posting to much or too fast”.
Happened 3x to me.
August 8th, 2008 at 11:40 pm
I don’t think bell has the right to control the services of their competitors or selectively choose what applications their members have a right to use at full speed. If they can not provide the speed the customer is paying for, they shouldn’t offer it as a service.
In my experience;
I was a bell member for 11 years and endured many outages and service interruptions, some lasted as long as 6 months. But I persevered and eventually my services were restored. My average bandwidth usage hovered around 2 Gigs per month. I’m in no way a power user or abuser of their services. My usage of the internet is sporadic at best and I was paying for a 5Mbit connection, so I expect to get 5Mbit when I need it.
Recently I began to use P2P to download Debian Linux and was shocked to discover I could not achieve a speed faster than 25KBs down. This discovery made me question Bell, they did not provide any answers as to why my traffic was slowed so much. This week I canceled my bell service in favor of another provider who was also offering a 5Mbs Service. I discover that this provider can not offer me the speed advertised on P2P, to make matters worse Bell decided to lower my bandwidth profile to 3 megs despite my providers request to give me 5Mbs of service.
So now I’m sitting on a 3Mbs profile that only allows me 1.6Mbs of speed (incorrectly configured), paying another provider for a 5Mbs service and downloading my Linux at only 25KBs or 200kbs, when I should be getting closer to 500KBs or 4000kbs. Furthermore I’m at the mercy of Bell to correct this profile issue that should never have taken place since my profile was already set correctly for a 5 meg service. The profile change took place on my official first day with my new provider and was reported to Bell by my ISP on Aug 7. What is supposed to take a maximum of 24 hours to correct has been scheduled by Bell to be corrected more than 60 hours after it was reported.
With Bells track record dealing with their competitors, according to DSL Reports and other similar websites, I expect to have many weeks of difficulty getting my profile set correctly to allow me the 5Mbs of speed I was enjoying at Bell 1 week ago. As for P2P, there are other ways to defeat Bells throttling mechanisms that I have since put into place. So Bell is now only harming their less savvy members internet experience.
Conclusions;
It’s unfair of Bell to deprive me of using the full speed of my connection, after all I’m paying for it.
To deprive me of my transfer speed when I change providers and to throttle my P2P connections with this new provider leads me to only one conclusion. Bell is trying to make my new provider appear as if they can not provide me the services offered. Therefore it’s obvious Bell is purposely sabotaging it’s competitors to prevent competition. According to numerous websites my situation is common among members who left Bell for a competitor.
Less savvy users would not know how to check the modems speed profile or recognize, despite the vendor, Bell ultimately controls the speed. So new members switching to the competition would conclude it’s the competitors fault the speed is so slow. This is clearly evident on websites like DSLreports.com.
From my perspective it appears Bell is playing games with the competition to ensure new users have a poor user experience and ultimately opt to return to Bell for a reliable service.
This opinion is based on my experiences with Bell and confirmed by my research on the web of the large number users who face the same issues that I am presently facing.