Availability zones protect against a failed building. They do not promise to outlast the region.
What happened
On 1 March 2026, AWS reported that objects struck a data center in the mec1-az2 zone of its UAE Region (me-central-1), creating sparks and fire; the fire department shut off power. A second UAE zone (mec1-az3) was impaired the next day, and AWS said it was not possible to launch new instances in the region. On 2 March, a zone in its Bahrain Region (me-south-1) lost power as well. On 3 March, AWS attributed the physical impacts in both regions to the ongoing conflict in the Middle East.
On 30 April 2026, AWS told customers that the Bahrain Region was unavailable and that they should recover resources in other Regions from remote backups. It advised UAE customers to migrate accessible resources and restore the rest from remote backups, and said recovery was expected to take several months.
On 15 September 2026, AWS said the Bahrain damage “spanned multiple Availability Zones and exceeded what our regional and multi-AZ services are designed to withstand.” It was unable to restore access to resources and data hosted exclusively in me-south-1. In the UAE, it could not restore resources and data hosted exclusively in mec1-az2.
The boundary
This is not a routine outage pattern, and it is not evidence that every region is at similar risk. AWS recommended migration after the first Bahrain zone was damaged in March and said most customers moved before the region became unavailable in April. It suspended relevant billing, continued recovery work in the other two UAE zones and plans a further Bahrain update in early 2027.
The provider did not publish a detailed post-incident report. The dated updates on the AWS Health Dashboard are the primary record.
The point
Only a copy held outside the affected region, and ideally outside the provider account, can be restored after that region is lost. Test the restore before you need it.
Sources
- AWS Health Dashboard: Middle East (Bahrain) Region availability
- AWS Health Dashboard: Middle East (UAE) Region availability
More incident dossiers
- Your spare zone needs room for your failure — Google Cloud
- A cloud told 495 customers: restore from your own backups — IDC Frontier (IDCF Cloud)
- Two gateways. One shared way to fail. — Microsoft Azure
- One failed AI request became 48 more — Microsoft Azure
- One suspended account cut off three clouds — Google Cloud / Railway
- Cloudflare went down, too — Cloudflare
- Azure’s global front door closed — Microsoft Azure
- One empty DNS record reached across AWS — AWS
All cloud outage evidence → · Test your cloud exit readiness →