• Lydia_K@lemmy.world
    link
    fedilink
    English
    arrow-up
    24
    ·
    20 hours ago

    The reality is that these multi region “clouds” are not nearly as resilient as they seem, and applications that have built on top of then even less so.

      • Lydia_K@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        9 hours ago

        There are at least two services I know of off the top of my head that rely on us-east-1, so even if you are using them in another region if us-east-1 goes down, so does that service in all regions (and so does your app)

    • gdog05@lemmy.world
      link
      fedilink
      English
      arrow-up
      12
      ·
      20 hours ago

      I’m far from an expert, but it seems like if one cloud node goes down, pay hackers to slam everything with bot traffic and the system is brought to its knees.

      • kablez@lemmy.world
        link
        fedilink
        English
        arrow-up
        14
        ·
        18 hours ago

        Correct. Plus it may not be as resilient as we think. Just because they’re billion dollar companies doesn’t mean they don’t follow the tradition of building systems with duct tape and a prayer.

        • Aceticon@lemmy.dbzer0.com
          link
          fedilink
          English
          arrow-up
          8
          ·
          12 hours ago

          In the age of MBA management, the removal of resilience such as fallback systems because “they’re doing nothing” is the norm.

          Nowadays Engineering stuff isn’t done according to Engineering Principles if it conflicts with short term profit maximization.

        • vaultdweller013@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          7
          ·
          18 hours ago

          Also even if the digital infrastructure is resilient, what about the physical infrastructure? A couple sticks of Nobels great invention and viola there’s now a gaping hole in your power transmission infrastructure or in the building itself if access can be achieved.

          • kablez@lemmy.world
            link
            fedilink
            English
            arrow-up
            9
            ·
            17 hours ago

            Not to mention the diminished teams who are now under resourced to handle their existing workloads. I’m sure they will do great as things spiral because the sloppy work of the AI has increased their workload, not reduced it. As the cracks begin to form on the human side of things, nothing could possibly go wrong. After all, the human element of any system is often the most secure. /s

            • vaultdweller013@sh.itjust.works
              link
              fedilink
              English
              arrow-up
              5
              ·
              14 hours ago

              You just triggered an old memory. Back when I was like 7 I remember watching some type of military debate on some type of public access thing, I think it was military historians and active duty guys talking about military communication over history interesting shit. Anyways something they say that’s kinda stuck with me was the worry of degrading autonomy within the military due to ever increasing ever improving communication capabilities and how it may get to the point that units become effectively headless should communications be disrupted.

              So what happens when leadership becomes reliant on AI tools and the data centers become inoperable? Because frankly that may be an actual worry sooner than later, or advantage depending on the longevity of the federal government.