IT has never been more popular than today 😅 The thing with IT: you don't notice it until it goes wrong.
The Crowdstrike outage is historical. Millions of workstations are crashing, around the world. Around 70% of Fortune 100 companies are affected. Big outages are almost never like this. Typically, it's a cloud service crashing, or a DNS problem, maybe a cable cut. But now it's millions of laptops and workstations crashing with a blue screen at the same time. The only comparisons to this would be malware outbreaks like Wannacry or Notpetya. With this incident, the software that was supposed to keep your machine up and running is now taking it down. End-point security software needs to run with low-level access in order to protect the computer. Unlike normal software, which can only crash itself, software with rights like this can crash the whole computer. To make matters worse, no other software is updated as frequently as security software. So, security companies have built very effective quality assurance processes, testing huge amount of software combinations to avoid crashes. Obviously, those processes failed here, as Crowdstrike's driver combined with their update channel configuration file crashed every machine. Maybe they tested one driver and shipped another - we don't know. Time will tell. Workarounds for the crash are known, but many of them are manual; you need to get a person physically at the keyboard of every crashed machine. Most critical systems are already up and running, but fixing every affected system will take days, maybe weeks. Image credit: Stella Qiu, Reuters