What do Glassdoor, iRobot, 1Password, Roku, WaPo, and Flickr have in common?
All intermittently went down last Wednesday, along with many other sites and apps.
The culprit? A multiple-hour outage affecting Amazon Web Services, the cloud infrastructure behind a good chunk of the internet. Major AWS clients include top Screen Time offenders like Netflix, Twitch, Reddit, Slack, and Spotify.
This outage affected one of 23 geographic AWS regions, but it still caused an influx of reports on DownDetector.
- Even puppies weren’t safe: “I got a cryptic message that an AWS outage is impacting my Petsafe feeder,” one user reported. “Do I have to somehow get home early so my animals have food?”
- Ironically, the issues prevented AWS itself from publishing updates to its Service Health Dashboard.
Issues stemmed from the Kinesis Data Streams API, according to Amazon, and efforts to increase system capacity.
Looking ahead: AWS outlined an updated game plan in response to the outage, including a move to larger CPU and memory servers. It’ll also direct more resources towards cellularization, a process that aims to isolate the effects of service failure.
Website of source