Elektrine lite

← Feed

@alienghic@timeloop.cafe

Post #3474223

2026-06-28 05:12 UTC

@doctormo@floss.social My current hack is I know there's some urls on my website that are dynamic enough the scrapers love searching them, but real people shouldn't be that interested. grep -E "suspicious|url|patterns" /var/log/apache2/access.log | cut -f 1 -d ' ' | asncounter --output-format tsv --top 4000 --no-asn -o /tmp/scum.tsv asncounter groups ip addresses by network so it restricts the trick of using a bunch of proxies on residential networks. Then I wrote a small program to select records out of scum.tsv and feed them to fail to ban like this, fail2ban-client set apache-noscript banip 117.188.0.0/20 1.80.0.0/13 180.72.0.0/14 ... Fail to ban will expire a ban after awhile, and I've got it set to increase the ban time for multiple hits. And then I kept doing that every morning, and some network bans kept getting increased. Just make sure to exclude your administrative ip address. And the AI companies cannot die soon enough.

Replies (1)

  • @yacc143@mastodon.social 2026-06-28 11:39

    @alienghic Makes me wonder when crawlers start packing reCAPTCHA-grade ML for this. Clean RL problem: agent walks a honeypot URL, IP/ASN gets banned, eventually the correlation surfaces and the policy learns to avoid traps. Guess not enough sites do it yet for the signal to bite. Genuine q: is your fail2ban purely local, or does it share ASN/CIDR bans via a community pool? If per-site, a crawler's done with you before the ban bites — kind of CrowdSec's pitch over fail2ban.

    Open ##3474224