• AWS says some resources hosted only in Bahrain or one affected UAE Availability Zone cannot be restored after conflict-related infrastructure damage
  • Bahrain damage crossed multiple Availability Zones, leaving customers without an in-region recovery path when the Region became unavailable

The fact

Amazon Web Services said on 15 September that it cannot restore access to some resources and data hosted exclusively in its Middle East (Bahrain) Region or in one affected Availability Zone in the United Arab Emirates. In Bahrain, AWS said physical damage crossed multiple Availability Zones and exceeded what its regional and multi-AZ services were designed to withstand. The company had advised customers to migrate critical workloads after earlier damage, and most had done so before another disruption made the Region unavailable in April.

In the UAE, resources and data held exclusively in the affected Availability Zone remain inaccessible, while recovery work continues elsewhere in the Region. AWS said most affected customers have re-established operations using backups or resources available outside the damaged infrastructure. The company plans further updates on the UAE recovery in the coming months and on Bahrain in early 2027.

The assessment

The Bahrain outage shows where redundancy inside one cloud Region stops. Spreading an application across Availability Zones can protect it against many local failures, but it does not provide a recovery route when damage affects several zones and the Region itself becomes unavailable.

Workloads that must survive that failure need usable copies elsewhere before the outage occurs. A secondary Region needs more than spare compute: data, credentials, network configuration and other dependencies must also be available and tested so the application can restart.

For BTW readers, the practical distinction is between redundancy inside a Region and recovery outside it. A multi-AZ design does not show whether a workload can survive regional loss; that depends on what has already been replicated, backed up and tested elsewhere.

What to watch

Watch AWS's next Bahrain and UAE updates for progress on restoring regional services. Customer post-incident reports on recovery times, off-region copies and dependencies that delayed failover will provide the clearest evidence of how different resilience designs performed when an entire Region became unavailable.