Amazon AWS: Some EC2 instances experienced connectivity issues and are now back to normal

On April 15th, Amazon AWS released an update saying that between 12:40 and 1:43 a.m. Pacific Time, a subset of EC2 instances located in a single Availability Zone (apne1-az4) in Asia Pacific Northeast Region 1 (AP-NORTHEAST-1) experienced connectivity issues. The issue originated from a simultaneous outage of primary and backup power, which affected the normal operation of the relevant EC2 instances. During this time, customers may experience higher error rates and latency on instances launched in the affected region, and other AWS APIs that rely in part on these EC2 instances may also be affected. Our engineers automatically stepped in within minutes and immediately began troubleshooting and implementing fixes. We do not expect this issue to occur again. Currently only a very small number of instances are still deployed on hardware affected by a power outage. Efforts to recover all affected instances and volumes will continue, but for immediate recovery, it is recommended to replace these unrecovered instances or storage volumes whenever possible. The issue has now been resolved and services are back up and running.