Widespread computer and internet outages are rare. They are often caused by errors in a centralized source, like a cloud service or server farm that users rely on. At times, damage to cables or other physical equipment can bring systems down.
But the outage that sent computers crashing around the world on Friday was unique — and meant that resolving the problem has been much harder and time-consuming. It started with faulty code pushed directly to PCs by CrowdStrike, a cybersecurity firm that says it’s used by more than half of Fortune 500 companies.
The faulty code caused the affected machines to enter an endless loop of reboots, taking them offline. That made it harder for the company to issue a fix online from afar.
Unless experts find a way to fix the machines remotely, each one will have to be fixed individually by a person manually removing the faulty code. CrowdStrike may also find a way to automate the process, which would speed things up.
