• hitmyspot@aussie.zone
    link
    fedilink
    arrow-up
    10
    ·
    26 days ago

    It seems odd that it’s all on one aws server cluster. Wouldn’t they all automatically activate their backup server on another instance?

    • VivianRixia@piefed.social
      link
      fedilink
      English
      arrow-up
      18
      arrow-down
      1
      ·
      26 days ago

      it’s us-east-1 that’s down, which is the one region with all the AWS services that aren’t easy to replicate, like DynamoDB, a major database service used by so so many companies out there. Its an all eggs in one basket kind of scenario. If it was any other region, we’d be fine.

    • queerlilhayseed@piefed.blahaj.zone
      link
      fedilink
      English
      arrow-up
      16
      ·
      26 days ago

      cross-region failovers are a thing, but they’re expensive to maintain so not everyone does it. I am kinda surprised one region failure had this much impact though

    • mistermodal@lemmy.ml
      link
      fedilink
      arrow-up
      4
      ·
      26 days ago

      You can get very fancy new alarm clocks that don’t need internet, skill issue. I’m convinced consumer grade IoT tech is some kind of irrational boomer exhibitionism thing

      • Scrath@lemmy.dbzer0.com
        link
        fedilink
        arrow-up
        2
        ·
        24 days ago

        If I host my own smart home with zigbee devices, does that still count as consumer IoT tech? The zigbee devices themselves are the typical philips hue, ikea tradfi, etc. devices

        • mistermodal@lemmy.ml
          link
          fedilink
          arrow-up
          1
          ·
          22 days ago

          No different kind of tech fetish (moving from public exhibitionism to being privately sealed in a latex suit)

  • Megaman_EXE@beehaw.org
    link
    fedilink
    arrow-up
    6
    ·
    26 days ago

    I feel like the internet should probably be hosted in as many different servers and physical locations as possible. It seems alarming that these things happen as often as they do

    • Uniquitous@beehaw.org
      link
      fedilink
      arrow-up
      6
      ·
      26 days ago

      It was only the one AWS region that went down, us-east-1. Unfortunately that’s the big kahuna. But they’ve got regions and availability zones around the globe. The services that went down due to the outage need to get some cross-region redundancy happening, but that’s the kind of thing that’s never a priority until the moment you need it.

      • Rooster326@programming.dev
        link
        fedilink
        arrow-up
        4
        arrow-down
        1
        ·
        26 days ago

        Even the places that did have redundancy - didn’t have actually tested failovers.

        Source: working at a place hosted in multiple locations. Didn’t mean shit. Somebody, if not many somebody’s, is definitely getting fired when the smoke clears.