Looking to better understand the recent traffic slowdowns experienced during interconnection feuds, M-Lab has released a new study (pdf) that analyzed transit and connection points between large last mile ISPs and transit operators such as Level3 and Cogent. While the report is clear not to affix specific blame for the sort of Netflix streaming issues seen by customers of Verizon and Comcast, they do clearly point out that the problems were the result of choices made by ISPs in their business relationships, and not congestion.
Both
Netflix and
Level3 have accused AT&T, Verizon and Comcast of intentionally leaving peering points un-upgraded to force content companies like Netflix to pay them for direct interconnection to bypass these intentionally congested links. These problems
evaporated once the deals were signed and ISPs were paid.
In contrast, AT&T, Verizon and Comcast have insisted that these are just run of the mill peering disputes, and have blamed the slowdowns on Netflix's choice of transit partners. ISPs often single out Cogent as the faulty party here given their history of feuds governing settlement-free peering arrangements.
The report however indicates that only customers on ISPs that were busy trying to get Netflix to pay direct interconnection fees experienced performance problems.
"In New York City, Access ISP Cablevision customers did not experience the same degradation symptoms as Comcast, Time Warner Cable, and Verizon customers when connecting across the Cogent network," states M-Lab. "Similarly, in Dallas and Los Angeles during the relevant time period, customers of Access ISP Cox did not experience a significant pattern or degree of degraded download throughput when connecting across Cogent sites."
Again, the study is quick to point out that because interconnection and peering agreements are confidential and much of the data is obscured from the public, they can't comfortably single out specific ISPs, even if fault is generally implied. You'll recall Verizon threatened to sue Netflix for even suggesting the ISP's choices were to blame for streaming performance degradation.
"We cannot tell whether any particular ISP between the user and a measurement point is “at fault,” what the contractual agreement between ISPs did or did not dictate vis-a-vis interconnection, or whether specific network modification was done to alleviate or magnify a given incident," states the study authors. "Similarly, we cannot identify the precise cause of performance problems (e.g. a broken router) in a path between a client on a given Access ISP and a M-Lab measurement point, although we take steps to narrow the range of possible causes."
Consumer advocates worry that as ISPs found themselves thwarted on neutrality fronts like throttling and usage caps, mega-carriers AT&T, Verizon and Comcast looked toward interconnection as the way to make the additional revenues they've always dreamed of. Because so much of this data is confidential, consumer advocates posit, mega-carriers likely felt comfortable in intentionally degrading traffic in order to get paid without it ever coming back to bite them.
The new M-Lab study goes a long way toward shedding a bright light on the culprit and potentially proving (or disproving) this narrative as fact by showing service degradations were a conscious business choice on the part of only some ISPs and not caused by normal network variables. The FCC is also conducting an investigation into these business disputes and has asked for much of the confidential agreement data shared only between companies. Combined with the M-LAB data, the agency may be able to ferret out the responsible culprits and shed a bright light on these disputes once and for all.
All of the study's data is
available here for those interested.