AWS Power Outage in Middle East Disrupts EC2 and Networking Services
On March 1, 2026, Amazon Web Services (AWS) experienced a significant power outage in its Middle East region (me-central-1), specifically affecting the mec1-az2 Availability Zone. The disruption was triggered by an unusual physical incident where external objects struck a data center, leading to sparks and a subsequent fire. This event necessitated a complete shutdown of power to the facility, including backup generators, as mandated by the fire department to manage the situation safely.
Impact on Services
The power loss had a profound impact on several AWS services within the affected zone:
– Amazon Elastic Compute Cloud (EC2): Instances became unavailable, disrupting applications and workloads running on these virtual servers.
– Amazon Elastic Block Store (EBS): Volumes attached to EC2 instances were rendered inaccessible, affecting data storage and retrieval operations.
– Amazon Relational Database Service (RDS): Databases hosted in the affected zone experienced downtime, impacting applications reliant on these databases.
Additionally, networking APIs faced significant throttling errors and failures. Customers reported issues with critical networking functions, including AllocateAddress, AssociateAddress, DescribeRouteTable, and DescribeNetworkInterfaces. These problems hindered the ability to manage network configurations and reassign Elastic IP addresses, complicating recovery efforts.
Timeline of the Incident
– 4:30 AM PST: The disruption began, with initial signs of connectivity and power issues.
– 4:51 AM PST: AWS officially acknowledged the problem and began investigating.
– 6:09 AM PST: AWS confirmed a localized power failure in mec1-az2.
– Afternoon: AWS deployed multiple configuration changes to mitigate API failures. By 2:28 PM PST, the AllocateAddress API showed signs of recovery.
– 6:01 PM PST: AWS confirmed the successful recovery of the AssociateAddress API, allowing customers to reassign Elastic IP addresses.
Mitigation and Recovery Efforts
AWS implemented several strategies to mitigate the impact:
– Traffic Routing: Requests were rerouted to unaffected Availability Zones within the region to distribute the load and minimize disruption.
– API Configuration Updates: Critical updates were deployed to restore functionality to networking APIs, enabling customers to manage their resources more effectively during the outage.
Despite these efforts, the physical infrastructure in mec1-az2 remained offline pending clearance from local authorities to safely restore power. AWS advised customers to operate out of alternate Availability Zones or Regions where possible and to restore data from their most recent EBS snapshots or backups.
Lessons and Recommendations
This incident underscores the importance of designing applications with redundancy across multiple Availability Zones to enhance resilience against localized failures. AWS emphasized that customers running redundant applications across multiple zones were largely insulated from the outage.
Organizations are encouraged to:
– Implement Multi-Zone Architectures: Distribute workloads across multiple Availability Zones to ensure high availability and fault tolerance.
– Regularly Backup Data: Maintain up-to-date backups to facilitate quick recovery in the event of an outage.
– Monitor Service Health: Utilize AWS Health Dashboard and other monitoring tools to stay informed about service statuses and receive timely updates during incidents.
As of the latest update, AWS continues to work on restoring full functionality to the affected Availability Zone and advises customers to follow recommended best practices to mitigate the impact of such incidents in the future.
