Widespread consequences stem from Amazon Web Services outage
BERKELEY, Calif. - AWS, Amazon Web Services, with 41% of the worldwide cloud-computing business – making it the world's largest, had a major failure Tuesday affecting tens if not hundreds of millions of people.
Many, major companies use its services to avoid building costly data server farms of their own. A year ago, while trying to add to their service capacity, they crashed the system taking down many websites.
According to DownDetector.com, AWS' problems, which appear to be the root of the overall outage problem, began about 9:30 a.m. Eastern Time. That impacted parts of North America, Europe and Asia. Social media lit up with complaints.
The outages are a who's who of major internet players including pretty much all Amazon services such as shopping, music, video, voice command systems such as Alexa, communications and its over arching web services serving many clients.
Other major Amazon clients, impacted: Xfinity, Google, Disney, Venmo, Capital One, Door Dash, Roku, Vanguard, Instacart, Netflix; four dozen in all.
A similar problem occurred in July took out many other websites for two hours," said Weaver."And there's only a few big providers and occasionally they make little mistakes and this seems to be a case of they made a minor mistake.
Back then, we spoke with UC Berkeley Computer Researcher Nicholas Weaver who said that Amazon screwed up part of its massive Web Service's computer system. "When you do an update, if you do a bad update, that takes down a lot of computers," said Weaver.
For example, Santa Clara University's physics department, caught up in this, was unable to allow students, just about to take exams, to access essential study materials.
"So, what we've done is simply reached out via traditional email and just gone ahead and posted exams and study guides and things like that via an email attachments. So, old school worked," said SCU physics department Professor Betty Young.
Amazon said all day long that there's no estimate when it will be fixed. Weaver likens this to Amazon locking its keys in the car. That's because the recovery tools needed to fix it are essentially locked in the trunk because they also must run on Amazon's system.