• socsa@piefed.social
    link
    fedilink
    English
    arrow-up
    32
    arrow-down
    4
    ·
    3 days ago

    A lot of people are getting this wrong. Cloudflare’s system is distributed the way it is to make DDoS attacks against individuals hosts and routes much more difficult. The goal is to block the traffic closer to the source to protect upstream routes. This morning it arguably worked as intended. The outage was around 30 minutes, and there was still intermittent connectivity during that time. Cloudflare didn’t collapse, it was struggling to separate legitimate traffic from DOS traffic, and throwing circui breakers and isolating nodes, because it was such a massive attack.

    • _cryptagion [he/him]@anarchist.nexus
      link
      fedilink
      English
      arrow-up
      19
      arrow-down
      1
      ·
      3 days ago

      except you’re getting it wrong, because it wasn’t an attack, according to Cloudflare. they fucked up a config file on one of their systems and that caused a cascading effect of failures in one system after another.

      it was quite literally not working as they intended.

    • wellheh@lemmy.sdf.org
      link
      fedilink
      arrow-up
      5
      ·
      3 days ago

      “30 minutes”- me when I lie on the internet. Where did you get that number? You realize we can check the news and see that big sites like x and chatgpt were down like 4 hours? Not only that, they said themselves it was not an attack but a misconfiguration. News were reporting it fixed around evening utc while the issue popped up around noon. That’s not a 30 min outage and is a huge failure.