Facebook scraper is back, though they don't seem to have been the bulk ot the excessive load. Apparently they own a whole IPv6 /29. The entire thing is blocked now at the iptables layer.

If anyone working at Facebook is having trouble legitimately using musl infrastructure, please inform your AI scraper department that they're the reason you can't access what you need and that they'll need to cease that activity and agree not to do it again in order to be unblocked.