Large Australian websites including Domain, Channel Nine, Foxtel Play and Domino's went down on Sunday afternoon after Amazon Web Service's Sydney zone suffered a two-hour power outage.
AWS first warned of the outage affecting Elastic Compute via its status page at 3:47pm, and an hour later confirmed the issue was related to a power problem. Sydney spent Sunday struggling through severe storms.
"We can confirm that instances have experienced a power event within a single availability zone in the AP-SOUTHEAST-2 region," it said.
"Error rates for the EC2 APIs have improved and launches of new EC2 instances are succeeding within the other availability zones in the region."
Our external vendor AWS confirmed Stan connection issues due to power outage within AWS. AWS & Stan are working to restore services ASAP.— Stan. (@StanAustralia) June 5, 2016
The outage also affected AWS services including ElastiCache, Redshift, Relational Database Service, Route 53 Private DNS, CloudFormation, CloudHSM, Database Migration Service, Elastic Beanstalk and Storage Gateway.
iTnews has contacted AWS for more detail. AWS has three availability zones in Sydney.
Power was restored to the affected facility around an hour and a half later and the company started working on restoring services.
AWS advised connectivity had been restored to the majority of Sydney region instances at around 9:30pm last night.
However it advised it was experiencing delays updating EC2 instance state data.
APIs were operating normally at the time but AWS spent the night recovering the remaining impacted instances.
"A couple of unexpected issues prevented our automated systems from recovering the remaining instances and volumes," AWS said.
"Most of these instances are hosted on hardware which was adversely affected by the loss of power.
"While we will continue to work to recover any affected instances or volumes, we recommend replacing any remaining affected instances or volumes if possible."
AWS was still restoring services early on Monday morning.