I also run a public cgit instance and get over 1M hits every day, although my pet projects are nowhere near the size or impact of kernel. I had to block (via nginx conf) cgit endpoints for diffs, blame, snapshots and historical commits, because nothing else works. Now they return 402 (payment required). I consider this my total defeat and it's killing me inside, but it is what it is.
As the article describes, it doesn't help, because the traffic originates from millions of unique residential IPs across hundreds of ASNs and countries.
Have you tried blocking a million IPs before? Fail2ban gets pretty shaky at even 200,000
The AI crawler traffic I’ve seen sends one request per ip and seemingly has an infinite pool of residential IPs. You can’t block the ASNs becuase you also block honest clients. IP blocks are the wrong solution.
And because I’m being negative I’ll also be constructive, IMHO the correct solution for fighting residential proxy crawlers is using RTT diffs this is one example https://github.com/Sakura-sx/Aroma
