• douglasg14b@programming.dev
    link
    fedilink
    arrow-up
    1
    ·
    edit-2
    6 days ago

    Yeah that’s mostly what I’m referring to.

    Backups are pretty easy, but service availability and failovers across cloud providers is stupid difficult. Not really from a compute standpoint but mostly from a data consistency/transactional standpoint.

    However, if you are using vendor specific services like AWS connect then you have to build and maintain multiple deep integrations into those services which effectively doubles your engineering efforts.

    • squaresinger@lemmy.world
      link
      fedilink
      arrow-up
      1
      ·
      6 days ago

      To save them from this desaster, backups would have been enough.

      Failover and service availability are nice, but for this use case not really necessary.

      If OOP had had regular backups to their PC and nothing else, it would have at least been pretty easy to rebuild everything on the same cloud provider and if they (understandably) didn’t want to trust AWS again, they’d at least have a basis for rebuilding on another provider’s cloud.