Amazon’s cloud computing arm, Amazon Web Services (AWS), temporarily shut down power at one of its United Arab Emirates data centers after “objects” struck the site, triggering sparks and a fire.
The incident comes amid heightened regional tensions following retaliatory strikes linked to the ongoing Iran conflict.
TL;DR
- Objects struck AWS Availability Zone mec1-az2 in the UAE at around 4:30 AM PST, causing sparks and a fire.
- Power was cut while firefighters contained the blaze, delaying connectivity restoration by several hours.
- Other Availability Zones in the ME-CENTRAL-1 Region continued operating, though customers faced API errors.
- AWS has not confirmed whether the incident is directly tied to regional missile or drone strikes.
AWS UAE Data Center Fire Disrupts ME-CENTRAL-1 Cloud Region
Amazon Web Services confirmed that at approximately 4:30 AM PST, one of its Availability Zones in the ME-CENTRAL-1 Region, identified as mec1-az2, was hit by objects that struck the data center, creating sparks and fire.
The UAE has been facing retaliatory missile and drone strikes from Iran following US and Israeli attacks. Airports, ports, and residential areas across the country and the broader Gulf have reportedly been impacted. However, when asked whether the data center incident was connected to the strikes, AWS did not confirm or deny a link.
“At around 4:30 AM PST, one of our Availability Zones (mec1-az2) was impacted by objects that struck the data center, creating sparks and fire,” AWS said.
An Availability Zone consists of one or more interconnected physical data centers that are isolated within a specific AWS Region. These zones are designed to operate independently to ensure resilience and high availability.
Following the incident, the fire department shut off power to the affected facility and its generators while crews worked to extinguish the blaze. AWS noted that it was awaiting permission to restore power and emphasized that restoring connectivity to the impacted zone would take several hours.
The company added that other Availability Zones in the UAE were functioning normally. Customers who had architected applications redundantly across multiple zones were not impacted by the event.
AWS Service Health Timeline Reveals Widespread API And Connectivity Issues
AWS’s official service health page detailed how the disruption unfolded across March 1:
-
Mar 01 4:51 AM PST: AWS began investigating service issues in the ME-CENTRAL-1 Region, citing connectivity disruptions affecting APIs and instances in mec1-az2.
-
Mar 01 6:09 AM PST: The company confirmed a localized power issue impacting a single Availability Zone. EC2 Instances, DB Instances, and EBS Volumes were unavailable in the affected zone, while other services experienced elevated error rates and latencies. Customers were advised to use alternate Availability Zones.
-
Mar 01 9:41 AM PST: AWS disclosed that objects had struck the data center, causing sparks and fire. Power had been cut by the fire department, and restoration was expected to take several hours. Customers running workloads redundantly across zones were not impacted.
-
Mar 01 12:14 PM PST: Customers reported errors in EC2 networking-related APIs, including AllocateAddress and AssociateAddress. AWS deployed configuration changes to mitigate some issues but noted that power restoration remained hours away.
-
Mar 01 2:28 PM PST: Recovery signs emerged for several EC2 APIs, though certain networking functions were still impaired. AWS continued mitigation efforts in parallel.
-
Mar 01 6:01 PM PST: AWS confirmed recovery of AssociateAddress API requests and introduced changes allowing customers to disassociate Elastic IP addresses from impacted resources. There was still no confirmed timeline for full power restoration.
-
Mar 01 10:46 PM PST: A localized power issue also affected another Availability Zone, mec1-az3. Customers experienced increased EC2 API errors across the region, and launching new instances was temporarily not possible. AWS projected recovery to be multiple hours away and advised customers to fail over to other Regions if possible.
Topics For More Insights
Middle East Conflict Raises Concerns For Regional Digital Infrastructure
The broader geopolitical backdrop has heightened concerns about the resilience of the Middle East’s expanding cloud and data center ecosystem.
Reports referenced strikes near major hubs including Dubai and Abu Dhabi airports, as well as landmarks such as the Burj Khalifa and Jebel Ali port. Neighboring cities such as Manama in Bahrain and Doha in Qatar also reported strikes.
While AWS has not directly attributed the data center fire to missile or drone debris, the timing has placed cloud operators and enterprise customers on alert. More details about the extent of physical damage and full recovery timelines are still awaited.
As regional tensions persist, the incident underscores the importance of multi-zone and multi-region redundancy strategies for mission-critical cloud workloads.
Join The Discussion