Turkey blocks BunnyCDN; Ctrl blog and 14 000 websites inaccessible

European Content Delivery Network (CDN) BunnyCDN is currently blocked in Turkey by major internet service providers in the country. The Turkish Information and Communication Technologies Authority’s (BTK) website confirms the blocking.

After technical analysis and legal consideration based on the law nr. 5651, administration measure has been taken for this website (b-cdn.net) according to decision nr. 490.05.01.2018.-369183 dated of the Information and Communication Technologies Authority.

b-cdn.net domain query, Turkish Information and Communication Technologies Authority (BTK)

BunnyCDN say on their blog that they’ve appealed the decision with the BTK. Turkey have blocked Wikipedia’s website in the country since , so I’m not convinced that anything will come of that appeal.

Update (): The ban have been lifted and BunnyCDN is once again accessible in Turkey. BunnyCDN haven’t provided any additional details on why they were blocked or whether they’d to make any changes at the behest of the Turkish regulators.

Turkey Blocks, a website dedicated to monitoring internet censorship in Turkey, confirmed in an email on Monday morning () that BunnyCDN (and Ctrl blog) is indeed blocked for internet users in Turkey.

The blocking is partially ineffective for some websites, including Ctrl blog, that have deployed custom-domains through Domain Name System (DNS) CNAME records. Anything that’s served directly from BunnyCDN’s primary delivery domain, b-cdn.net, is inaccessible, however. This is an interesting difference as it shows that the ISP’s understanding of the inner workings of their own DNS server software may be incomplete.

Turkey Blocks were also able to confirm the presence of deep-package inspection (DPI) that looks at the content of users’ data to intercept the HTTP Host header on unencrypted channels as well as server name indication (SNI) domain names in encrypted traffic.

, I wrote about how the web have become too centralized and why we need the distributed web. It all felt very idealistic and theoretical at the time, but today it feels a lot more up close and personal.

I’ve been tinkering with creating a mirror of this website to be distributed through distributed content networks like Dat and InterPlanetary File System (IPFS). These would possibly have been more resistant to internet censorship. This isn’t quite ready yet, though parts of Ctrl blog are now indeed accessible over these networks. More on this later.

In the meantime: Copies of every article from this website are available through the Internet Archive. The newest articles may not yet be available and they may not all be up to date though I’ve pushed new copies of everything this morning.