Is the daily down time of lemmy.world because of attacks, system instability, or some kind of resource issue?

Comments

• freamon@endlesstalk.org ⁨1⁩ ⁨year⁩ ago

  Sync was released recently, and - like many apps before it - directed its users to lemmy.world as a ‘default’ instance, so they’ve had an influx of users to contend with.

  Also, if you ask for the ‘next page’ of communities via their API, it’ll just keep feeding you the same ones, over and over, even if you ask for Page 1 Billion, so there’s probably some bots, crawlers, front-ends etc thrashing the hell out of it.

  I don’t know about DDoS attacks, but while they’re happening, it seems to act a catch-all to blame any problems on.

  source

  • Rooki@lemmy.world ⁨1⁩ ⁨year⁩ ago

    Its not the new user influx. If all users would magically disappear it will still be down. Its an issue with the database and the queries from lemmy.

    source
• Rooki@lemmy.world ⁨1⁩ ⁨year⁩ ago

  Its a lemmy bug with the backend. The Database breaks down because of some queries and is stuck until a restart.

  source

  • madcaesar@lemmy.world ⁨1⁩ ⁨year⁩ ago

    I know scaling DBs can be tricky, but I also know there are bootstrap solutions that go pretty high up before needing custom work.

    I don’t know enough about the lemmy infrastructure, but did they build some custom thing scratch framework or did they start with something stable and tested?

    source

    • Rooki@lemmy.world ⁨1⁩ ⁨year⁩ ago

      The thing is, it is not the amount available databases. Rather 1 query takes super long and it blocks everything

      source
    • 31337@sh.itjust.works ⁨1⁩ ⁨year⁩ ago

      Lemmy uses the Diesel ORM. Lemmy uses a large collection of Rust libraries, so I guess you could say they rolled their own framework. I’ve never encountered a framework that I believe could handle non-trivial high-traffic web applications. I worked on a project that used Django for years. By the time we were done, we bypassed almost all of Django’s functionality to get it to scale with our data and users.

      source
    • MrPoopyButthole@lemmy.world ⁨1⁩ ⁨year⁩ ago

      Its Rust and Postgres

      source
  • MrPoopyButthole@lemmy.world ⁨1⁩ ⁨year⁩ ago

    Do you have the link to the Github issue for that?

    source

    • Rooki@lemmy.world ⁨1⁩ ⁨year⁩ ago

      No as the admins have to identify the exact troublesome db query themselves.

      source

      • -> View More Comments
• DudeWTF@lemmynsfw.com ⁨1⁩ ⁨year⁩ ago

  Might be the new admin call out for volunteers from a few days ago. If so, I think someone just failed the first day exam. The only way to deal with this is far far far more transparency about ineptitude, and someone that learns extremely quickly.

  source

  • AFKBRBChocolate@lemmy.ca ⁨1⁩ ⁨year⁩ ago

    It’s been going on since well before that, and they’ve been pretty good at transparency generally, so that seems unlikely to me.

    source