• Australis13
    link
    fedilink
    1484 months ago

    This is why you do staged rollouts of updates… not the entire planet at once.

      • shameless
        link
        fedilink
        English
        214 months ago

        So true, this really highlights the risk of updates impacting critical systems vs critical systems being exposed to critical vulnerabilities. Its a real balancing act.

        • @[email protected]
          link
          fedilink
          English
          14 months ago

          It actually highlights the risk of having unaudited third party software running on your critical infrastructure

      • @[email protected]
        link
        fedilink
        English
        114 months ago

        I don’t know exactly how crowd strike works, but this sounded like a “virus signatures” update (IE not a software update per se). And thats what caused the issue.

        I think “real time virus protection” is why people use it so they expect the signatures to get updated asap/with little to no human intervention.

        This is a crowd strike epic fail…for how they let their software blue screen systems with a virus signature update.

        • sunzu
          link
          fedilink
          34 months ago

          Can somebody explain why the down votes?

          • @[email protected]
            link
            fedilink
            English
            154 months ago

            Mostly because it’s simply not that easy. Devs go where support is at and follow market share (2000s era Mac gamer memes.)

            If you look at the Linux community as a whole it’s a wasteland of competing parties and standards. So it’s not developing for linux it’s developing for distros^hardware.

            Windows is shit and it’s pretty well known that it’s getting worse… but it’s still the standard and unfortunately until Linux starts unifying and becoming more stable for developers it’s unlikely to become more compelling for the broader market to switch to.

            TLDR; every time a new conflict breaks out hop in that thread and say “give peace a chance” and see how well that gets received.

            • @[email protected]
              link
              fedilink
              English
              64 months ago

              Windows is actually steadily improving from a security point of view. MS is finally starting to deprecate ancient garbage like NTLM, UWP apps are sandboxed and there’s even talk of rewriting core libraries in Rust to make them memory safe.

              • @[email protected]
                link
                fedilink
                English
                14 months ago

                This is true enough. In general though I think it’s finding a tipping point between investors and having a good OS. Per the usual every other OS pattern they follow it probably will be on the struggle bus until the next version.

            • sunzu
              link
              fedilink
              2
              edit-2
              4 months ago

              Are we talking linux deskop for usage at end points?

              Seems odd, we CAN run servers on it but end points can’t be done properly.

              I don’t know shit about shit but linux desktop was a pleasant surprise as a gamer.

              • @[email protected]
                link
                fedilink
                English
                44 months ago

                Steams been a massive contributor to that.

                It’s a demonstration that if we focus on a common goal that Linux development can actually be pushed forwards. So this is definitely an improvement for end users - and I expect it will improve in the future… But broadly speaking there are too many requirements for some level of troubleshooting knowledge.

              • @[email protected]
                link
                fedilink
                English
                3
                edit-2
                4 months ago

                That comment you answered to is full of shit, desktop Linux works fine for many companies. And no dev ever chooses Windows lol

            • @[email protected]
              link
              fedilink
              English
              14 months ago

              A wasteland of competing standards lol

              More like a bountiful harvest. Even the dogshit programs windows users buy are mostly made with FOSS libraries.

              The real answer is windows apologists don’t want to hear Linux users gloating about how this never happened to linux, and how dogshit their beloved os is

          • @[email protected]
            link
            fedilink
            English
            84 months ago

            Because “just run Linux lol” is unrealistic and naive, it gets said on every thread, and it gets incredibly tiring.

            • sunzu
              link
              fedilink
              14 months ago

              i guess running Windows PC is realistic and smart lol

              • @[email protected]
                link
                fedilink
                English
                64 months ago

                When I have a very large, very expensive piece of manufacturing equipment whose control software only runs on Windows, yes, it is.

                • sunzu
                  link
                  fedilink
                  24 months ago

                  specific use case requires specific set up…

    • @[email protected]OP
      link
      fedilink
      English
      1104 months ago

      It’s basically corporate anti-virus software. Intended to detect and prevent malware.

      • Alimentar
        link
        fedilink
        English
        314 months ago

        Apparently it’s the next iteration of AI based antivirus where it uses smart algorithms to detect system behaviours and makes assessments on whether they’re malicious or not

        • Boxscape
          link
          fedilink
          English
          364 months ago

          Apparently it’s the next iteration of AI based antivirus

          CrowdskyStrikenet

        • sukotai
          link
          fedilink
          English
          114 months ago

          obviously, A.I consider microsoft as a malicious software. Sometimes, A.I is very accurate 😁

      • @[email protected]
        link
        fedilink
        English
        14 months ago

        Can you tell whether this update was delivered by Crowdstrike’s own update delivery pipeline of via Window’s update pipeline?

    • @send_me_your_ink
      link
      English
      44 months ago

      It’s software put on every machine so that the company can quickly isolate it if/when something bad happens (or it falls out of security compliance). To do this is requires a constant Internet connection, insanely high privileges on the machine and frequent updates to be appraised of risks.

      That risk update went off the rails and into the next state.

  • @[email protected]
    link
    fedilink
    English
    624 months ago

    I work in QA on the night shift at a video game company. It was absolute chaos at work tonight lmao we only had a grand total of 6 working PCs between all of us

  • @[email protected]
    link
    fedilink
    English
    57
    edit-2
    4 months ago

    Company spyware. We have that on our devices. They used to have an “about” stored locally on the app, but removed it and a web connection is required to view the docs. Basically says it downloads/sees everything on your device and checks for threats. Thing is a few people have been fired for having things in their devices they shouldn’t. I didn’t ask what it was, nor did I hear how these things were “threats”, but nonetheless they were fired. Too many people treat company hardware like “free device, bro!” and put all sorts of personal stuff on the device. Most industries it’s probably not too big of a deal, but for mine if there’s an incident that happens when you were busy watching Netflix or something instead of doing your job you’re fucked. First thing they’ll do is check your device and crowdstrike to see what you were doing, and even if you weren’t watching Netflix all your personal data will be exposed.

    • @[email protected]
      link
      fedilink
      English
      204 months ago

      They definitely could, but most cybersecurity departments are paid too much to worry about minor items like that. If HR tells us to look into a specific user and gets the proper approvals so that everything is in compliance, we’ll definitely get someone on the team to do it, but otherwise if we happen to see evidence of unapproved usage, we’re mostly going to overlook it unless it could lead to something dangerous to your machine or the company as a whole.

      EDRs like Crowdstrike can see very very nearly everything you do though, definitely everything you would care about.

  • @[email protected]
    link
    fedilink
    English
    474 months ago

    Yikes. I feel sorry for all the help desk and support staff that has to deal with this chaotic mess all day.

    • @[email protected]
      link
      fedilink
      English
      514 months ago

      Won’t take that long, security researchers are already decompiling the update to see if it was malicious or incompetence.

      • @[email protected]
        link
        fedilink
        English
        134 months ago

        You won’t find the incompetence in the software no matter what.

        If you fail to assume that the software contains issues – if you fail to understand that your software is made by humans and humans make mistakes, not because they’re bad but because they’re human – and if you fail to implement mechanisms to feel gracefully with inevitable failures, THAT is the incompetence.

        Failures are systemic.

        • @[email protected]
          link
          fedilink
          English
          324 months ago

          Oh yes I make those failures myself, testing and staging and limited release schedules save my human failures from breaking the world

          • @[email protected]
            link
            fedilink
            English
            34 months ago

            One funny thing about humans is that they aren’t just gloriously fallible: they also get quite upset when that’s pointed out. :)

            Unfortunately, that’s also how you end up with blameful company cultures that actively make reliability worse, because then your humans make just the same amounts of mistakes, but they hide them – and you never get a chance to evolve your systems with the safeguards that would have prevented these.

      • @[email protected]
        link
        fedilink
        English
        24 months ago

        Think big. This may have had a target. But hitting the target only wasn’t possible so everyone got hit.

        It’s possible those responsible only had this weapon that was capable of hitting the target, maybe the plan was to disrupt world flights to make someone late tomorrow, who knows. Maybe poo-tin or Xi-the-Pooh wanted to hit America and its allies?

        • @[email protected]
          link
          fedilink
          English
          5
          edit-2
          4 months ago

          A state actor will use more precise techniques to attack specific targets. Think SolarWinds and Stuxnet.

          Ransomware doesn’t apply here and tends to depend on phishing first anyway.

          Even terrorists have specific targets in mind.

          So it’s either Bond villains or incompetence.

          Edit: The only way i can fit your comment would be an incompetent script kiddy. Even then, doesn’t make sense as all systems were not directly attacked, as would be the case, but rather through what would have to be a side-channel attack, so no.

    • @[email protected]
      link
      fedilink
      English
      154 months ago

      Never attribute to maliciousness that which can be explained by incompetence.

      That said, I’m sure the Crowdstrike CEO is currently on a phone call with three of their pet Congresscritters asking if they can get a $100M grant to harden their systems against Russia/China/NKorea/Antifa interference right now.

      • @[email protected]
        link
        fedilink
        English
        34 months ago

        If I ever become a super 1337 hacker I’m going to setup all of my exploits to look like it could be regular mismanagement, thanks for the advice

      • @[email protected]
        link
        fedilink
        English
        24 months ago

        While being simultaneously gang…handled by the unnamed 3-letter agencies representatives