At some point I have to start wondering if Putin pays these sorts of people.

  • db0@lemmy.dbzer0.comM
    link
    fedilink
    arrow-up
    10
    ·
    21 days ago

    The DB is all that matters. Python can scale very well through parallelization. So long as one doesn’t restrict themselves to one process, there’s really little chokepoint.

    • ad_on_is@lemm.ee
      link
      fedilink
      English
      arrow-up
      1
      ·
      20 days ago

      Nope… CPU and memory usage matter as well… if they get exhausted, you get throttling. This also has an impact on server-costs… Why run 2 instances of something that serves 4k requests/second over one instance that serves 9k/s (just an over-exaggerating example)

      • db0@lemmy.dbzer0.comM
        link
        fedilink
        arrow-up
        4
        ·
        edit-2
        20 days ago

        That’s why I say if you don’t restrict to single process. As to why something which might be slightly more inefficient (it’s not going to be that much), it’s because of ease of development and pool of potential developers to help you with it.