AWS Suffers Another Cloud Outage
AWS EC2 and S3 taken down in US regions as yellow-tick horror shocks GitHub
Amazon Web Services (AWS) has suffered a large cloud outage which affected its Elastic Compute Cloud (EC2) and S3 platforms.
AWS notified customers of “elevated error rates” for these services in the North Virginia region, as well as US-STANDARD region for S3.
Elevated errors
At 12:36 AM PDT, AWS said it was investigating elevated errors for requests made to Amazon S3 in the US-STANDARD Region. Then at 1AM PDT, it said it was working to determine root cause of the elevated error rates and latencies for Amazon S3 in the US-STANDARD Region.
An update came eight minutes later as AWS claimed to have identified the cause of the errors and latencies for requests to the US-STANDARD region.
At 1:52AM PDT, Amazon said: “We are actively working on the recovery process, focusing on multiple steps in parallel. While we are in recovery, customers will continue to see elevated error rate and latencies.”
The GitHub status Twitter account alerted users to the outage this morning, saying that it had identified an issue with its provider and was monitoring the situation.
Some AWS users questioned the specific meaning of the “elevated error rates” message, speculating that by using this terminology, AWS can get away with not breaching its SLAs.
“They should worry about that a lot. Amazon are notoriously bad at communicating during outages. They’ve gotten better, but they’re big enough that it should have priority,” said user mrsirduke writing on a thread discussing the outage on ycombinator.com.
Panic then ensued as AWS’ green ticks on its status page, which indicate healthy service, started to turn yellow.
“The icon got yellow. I repeat: YELLOW,” exclaimed user kore_sar.
“It only goes red when a nuclear event occurs, obliterating most of humanity and only the machines remain,” replied user hughstephens.
TechWeekEurope has asked AWS for a statement on the outage, but has yet to receive a reply.
Cloud application platform Heroku was one of the customers affected. This morning the company said: “Heroku engineers have confirmed issues with an upstream provider. At the moment apps may not build successfully, customers may attempt to retry app builds for now while we work on a solution.”
Amazon Web Services’ last hiccup was in July, when connectivity issues from an “external Internet service provider” affected traffic from some end-user networks and took out sites such as Pinterest and Netflix. AWS said its services were not affected.
“We are currently monitoring an external Internet provider issue that is causing interrupted service connectivity to AWS services for some customers,” said the firm. “AWS services are not affected and continue to operate normally.” The problems lasted for around 40 minutes.
