You seem to be implying that you don't believe/can't see the GFW

No, that's not what I meant. I thought mandatory content filtering at the border means traffic throttling at the border, deliberately or accidentally rate-limiting the traffic, now 

I think he was referring to GFW and the side effect of deep packet inspection. 

In fact, we designed a small experiment to locate the hops with GFW presence, and then try to match them with the bottleneck hops. We found only in 34.45% of the cases, the GFW hops match the bottleneck hops. 

Best,
Pengxiong Zhu
Department of Computer Science and Engineering
University of California, Riverside

On Mon, Mar 2, 2020 at 1:13 PM Matt Corallo <nanog@as397444.net> wrote:

> find out direct evidence of mandatory content filtering at the border

You seem to be implying that you don't believe/can't see the GFW, which
seems surprising. I've personally had issues with traffic crossing it
getting RST'd (luckily I was fortunate enough to cross through a GFW
instance which was easy to avoid with a simple iptables DROP), but its
also one of the most well-studied bits of opaque internet censorship
gear in the world. I'm not sure how you could possibly miss it.

Matt

On 3/2/20 2:55 PM, Pengxiong Zhu wrote:
> Yes, we agree. The poor transnational Internet performance effectively
> puts any foreign business that does not have a physical presence (i.e.,
> servers) in China at a disadvantage.
> The challenge is to find out direct evidence to prove mandatory content
> filtering at the border, if the government is actually doing it.
>
> Best,
> Pengxiong Zhu
> Department of Computer Science and Engineering
> University of California, Riverside
>
>
> On Mon, Mar 2, 2020 at 8:38 AM Matt Corallo <nanog@as397444.net
> <mailto:nanog@as397444.net>> wrote:
>
>     It also gives local competitors a leg up by helping domestic apps
>     perform better simply by being hosted domestically (or making
>     foreign players host inside China).
>
>>     On Mar 2, 2020, at 11:27, Ben Cannon <ben@6by7.net
>>     <mailto:ben@6by7.net>> wrote:
>>
>>     
>>     It’s the Government doing mandatory content filtering at the
>>     border.  Their hardware is either deliberately or accidentally
>>     poor-performing.
>>
>>     I believe providing limited and throttled external connectivity
>>     may be deliberate; think of how that curtails for one thing;
>>     streaming video? 
>>
>>     -Ben.
>>
>>     -Ben Cannon
>>     CEO 6x7 Networks & 6x7 Telecom, LLC 
>>     ben@6by7.net <mailto:ben@6by7.net>
>>
>>
>>
>>>     On Mar 1, 2020, at 9:00 PM, Pengxiong Zhu <pzhu011@ucr.edu
>>>     <mailto:pzhu011@ucr.edu>> wrote:
>>>
>>>     Hi all,
>>>
>>>     We are a group of researchers at University of California,
>>>     Riverside who have been working on measuring the transnational
>>>     network performance (and have previously asked questions on the
>>>     mailing list). Our work has now led to a publication in
>>>     Sigmetrics 2020 and we are eager to share some
>>>     interesting findings. 
>>>
>>>     We find China's transnational networks have extremely poor
>>>     performance when accessing foreign sites, where the throughput is
>>>     often persistently
>>>     low (e.g., for the majority of the daytime). Compared to other
>>>     countries we measured including both developed and developing,
>>>     China's transnational network performance is among the worst
>>>     (comparable and even worse than some African countries).
>>>
>>>     Measuring from more than 400 pairs of mainland China and foreign
>>>     nodes over more than 53 days, our result shows when data
>>>     transferring from foreign nodes to China, 79% of measured
>>>     connections has throughput lower than the 1Mbps, sometimes it is
>>>     even much lower. The slow speed occurs only during certain times
>>>     and forms a diurnal pattern that resembles congestion
>>>     (irrespective of network protocol and content), please see the
>>>     following figure. The diurnal pattern is fairly stable, 80% to
>>>     95% of the transnational connections have a less than 3 hours
>>>     standard deviation of the slowdown hours each day over the entire
>>>     duration. However, the speed rises up from 1Mbps to 4Mbps in
>>>     about half an hour.
>>>
>>>
>>>     We are able to confirm that high packet loss rates and delays are
>>>     incurred in the foreign-to-China direction only. Moreover, the
>>>     end-to-end loss rate could rise up to 40% during the slow period,
>>>     with ~15% on average.
>>>
>>>     There are a few things noteworthy regarding the phenomenon. First
>>>     of all, all traffic types are treated equally, HTTP(S), VPN,
>>>     etc., which means it is discriminating or differentiating any
>>>     specific kinds of traffic. Second, we found for 71% of
>>>     connections, the bottleneck is located inside China (the second
>>>     hop after entering China or further), which means that it is
>>>     mostly unrelated to the transnational link itself (e.g.,
>>>     submarine cable). Yet we never observed any such domestic traffic
>>>     slowdowns within China.
>>>     Assuming this is due to congestion, it is unclear why the
>>>     infrastructures within China that handles transnational traffic
>>>     is not even capable to handle the capacity of transnational
>>>     links, e.g., submarine cable, which maybe the most expensive
>>>     investment themselves.
>>>
>>>     Here is the link to our paper:
>>>     https://www.cs.ucr.edu/~zhiyunq/pub/sigmetrics20_slowdown.pdf
>>>
>>>     We appreciate any comments or feedback. 
>>>     --
>>>
>>>     Best,
>>>     Pengxiong Zhu
>>>     Department of Computer Science and Engineering
>>>     University of California, Riverside
>>