AWS reports fire after objects hit the UAE data centre, one Availability Zone offline

Amazon Web Services said a fire broke out at one of its data centres in the United Arab Emirates on 1 March 2026 after “objects” struck the facility, prompting emergency crews to cut power and take the affected site offline. According to the company, the incident was contained to a single Availability Zone in the Middle East, UAE Region, and other zones continued to operate normally. AWS added that restoration work would proceed once authorities cleared a safe power up, and that recovery could take several hours.

Impact on services and customer guidance

AWS said the strike created sparks and a fire, after which the fire department shut down utility power and generators to enable firefighting. The company noted that redundancy across Availability Zones is intended to keep regional services running even when one zone faces a disruption. Customers who architect workloads across multiple zones were advised to fail over to healthy zones while teams worked to restore the impacted site.

As of 2 March 2026, AWS said it would bring the affected infrastructure back only after safety checks, power stabilisation and network validation. Customers running single zone deployments may see longer recovery times, whereas multi zone and disaster recovery setups are expected to reduce downtime to business acceptable levels.

Context and what we know so far

The company did not specify the nature of the “objects” or attribute the incident to any actor. It also declined to comment on whether the strike was linked to the wider regional tensions reported over the weekend. The affected zone in the UAE was identified in company communications as mec1-az2, part of the me-central-1 region. AWS has not reported damage beyond the site that was powered down for safety, and said other zones in the country remained available for customers.

Amazon, led by CEO Andy Jassy, derives a significant share of operating profit from its cloud unit. AWS, which is headed by CEO Matt Garman, positions its global infrastructure around fault isolation, meaning Availability Zones are physically separated with independent power, cooling and networking. Industry analysts said that architecture is designed to limit the blast radius of incidents like fires, localised floods or fibre cuts, thereby allowing customers to continue operations by shifting traffic to unaffected zones.

How should AWS customers in India and the Gulf respond

• Confirm whether any workloads are pinned to the affected UAE zone, then initiate failover to another zone in the same region as per runbooks. Where possible, place standby capacity to absorb traffic spikes.
• Validate backup and restore pipelines, including snapshots, database replicas and object storage versioning, and verify that identity and access, secrets and networking policies are pre-provisioned in alternate zones.
• Exercise cross region recovery for critical applications that serve users across West Asia and India, accounting for latency, compliance and data residency needs.
• Enhance observability so teams can quickly distinguish infrastructure events from application defects, improving meantime to detect and recover.

Why this matters for India’s digital economy

Enterprises, startups and government programmes across India increasingly rely on cloud regions in the Middle East for latency sensitive workloads that serve the Gulf diaspora, travellers and cross border trade. A zonal incident, even when contained, can ripple into customer experience and service level metrics if applications are not designed for resilience. Founders and CTOs in India often pair the UAE region with deployments in India and Europe to balance performance with continuity, a practice that becomes more relevant in light of this outage.

According to the company, teams on the ground are coordinating with emergency responders and utility providers to restore the facility safely.