Say a word about poor twitter
An attempt by the ILV to “slow down” communication with Twitter resulted in local “Internet cataclysms”.
Unfortunately, any useful technical information on how exactly ILV decided to implement what what publicly stated, not so much. You can even say that it practically does not exist. So it remains only to speculate on the observations and available data.
From the observations at the moment, we have: scattered “local” messages that twitter really opens slowly (or does not open at all). Moreover, it seems that most of them are related to the mobile application.
At the same time, at the same time, even through Rostelecom (by wire), Twitter itself opens quickly, and so does the video from its CDN.
However, some of our tests show that some providers and indeed there is a picture of a sharp limitation of the download speed, if in SNI the domain was specified in which contains (sic!) textt.co… Exactly contains… Those. microsof is also included in the samplet.com, and githubusercontent.com (the domain where this social network keeps user avatars and all sorts of “little things”) and so on. And this is reproduced even if you send requests to the IP of your own server (that is, it is not at all necessary – of Twitter itself).
Why the implementation is so clumsy is a huge question:
On the one hand, it looks either as a profanation on the part of the entire department as a whole, or as sabotage on the ground, and on the other, it may well turn out to be a typical problem that the executor’s programmers, having no idea what the customer wants, made the tool as illogical as possible and inconvenient from the point of view of the customer. Here I will remind you once again that, in fact, there is no technical information on how exactly the RKN implements this very slowdown in the public domain. And the possible implementation options are too many to carry out analytics for each. So, at the moment, our working hypothesis is the use of software and hardware systems (the very ones that providers were obliged to install according to the 90-FZ (on the sovereign Internet)), however, as far as I remember personally, until recently, providers unanimously stated that that there is still no such ready-made “piece of iron”.
Another candidate for this role is SORM-3, but if this particular complex is used, a natural question arises as to why the means of operational-search measures are suddenly used to shape the traffic of social networks.
Well, the third candidate is the theory of local “agreements” with specific providers (fortunately, there are in fact only two large federal ones: Rostelecom and ER-Telecom. The rest are either bought by one of them, or are “too small”.
Another funny coincidence is that very soon after the RKN’s statement, a large number of government agencies’ websites suddenly fell.
Most likely, this is just a coincidence, but if you look from the theoretical point of view, “could they have messed up so that such an effect was obtained”, my answer as a technical specialist is yes, they could (especially since everyone, as usual, denies everything). True, for this you need to be absolutely “poor”. However, despite the fact that the SNI detector “clings” to sites like microsoft.com, doubts remain less and less.
But, as I said several times already, without any factual information, all that remains is to fantasize and speculate. Well, if you want to – weave conspiracy theories.
For example, here, just somewhere at the same time, one of the OVH data centers caught fire. Coincidence? 🙂
By the way, in fact, the above proposal is not as joke as it might seem at first glance.
The scale of the defeat at OVH is quite large: one data center burned down, part of the second was damaged, the third was “in the process of the fight between firefighters and fire”, and only the fourth block was not touched.
Considering that OVH is one of the largest hosters not only in Europe, but also in the world, the restructuring of global routing in connection with the fire could really hit the “working of the Internet” well from the point of view of the layman.
