Several Web services -- including Netflix, Instagram, Vine and Airbnb -- were knocked out of action last night by a glitch at one of Amazon's data centres.
The problems started late Sunday evening, the BBC reports. All of the above services rely on Amazon's cloud-based network.
Cloud software company Heroku and Web automation service IFTTT were also affected.
The software problems were traced to a data centre in northern Virginia, USA, that couldn't keep up with demand. It all went Pete Tong at about 9pm UK time, 4pm Eastern time. For several hours, access to the above services was intermittent, with people reporting problems logging in, and that pages were taking an age to load.
Instagram was the first to acknowledge the issue, and put out a message on Twitter. The others followed soon after. They're all back up and running now as normal.
The glitch affected databases, software that spreads queries across servers, and the code controlling the core computers that underlie the entire service. Amazon said it was caused by the "partial failure of a networking device".
It's not been a good week for Amazon. Just a few days earlier, its US store went down for about half an hour. Apple's cloud services suffered an outage last week, too.
It is a bit worrying, that a single glitch in one data centre can bring all these services to a standstill. Just over a year ago, Apple co-founder Steve Wozniak came out as a critic of the cloud. "I really worry about everything going to the cloud," he said. "I think it's going to be horrendous. I think there are going to be a lot of horrible problems in the next five years." Maybe he's got a point.
Were you affected by the glitch? Do you think we put too much trust in the cloud? Let me know in the comments, or on our Facebook page.