AWS said on September 15 that it cannot restore access to resources and data hosted exclusively in Bahrain’s ME-SOUTH-1 Region or in the UAE’s mec1-az2 Availability Zone. The update came after months of recovery work following conflict-related physical damage to AWS facilities, and it draws a sharp line between data replicated elsewhere and data kept only in the affected locations.
AWS also said recovery continues in the UAE’s mec1-az1 and mec1-az3 zones, where it is replacing damaged infrastructure. Most customers, the company said, re-established operations using backups or by copying data that remained accessible in other Regions.
AWS’s recovery scope in Bahrain and the UAE
| Location | AWS Region or zone | Resources and data AWS cannot restore | Current status | Customer recovery path |
| Bahrain | ME-SOUTH-1 Region | Resources and data hosted exclusively in the Bahrain Region | Damage spans multiple Availability Zones; AWS plans another Bahrain update in early 2027 | Backups, accessible data and operations in other Regions |
| United Arab Emirates | ME-CENTRAL-1, mec1-az2 | Resources and data hosted exclusively in mec1-az2 | Recovery continues in mec1-az1 and mec1-az3, with infrastructure replacement under way | Backups, accessible data and operations in other Regions |
An AWS Availability Zone is a cluster of one or more data centers within a geographic Region. Cloud architectures commonly spread workloads across zones so that a failure in one location does not take down the whole service. That model has a hard boundary, however: it is designed around particular failure assumptions, not every possible form of physical destruction.
AWS says the Bahrain damage crossed several Availability Zones and exceeded what its regional and multi-AZ services were designed to withstand. In the UAE, AWS’s March 2 update described direct strikes on two facilities. Structural damage, disrupted power delivery and water damage from fire-suppression activity complicated the recovery.
What geographic redundancy could—and could not—handle
The UAE incident shows the difference between a single-zone failure and damage affecting several facilities in the same Region. AWS said one UAE zone, mec1-az2, now contains resources and data that cannot be restored when they were hosted exclusively there. Work continues in the other two named zones, mec1-az1 and mec1-az3.
Bahrain presents the broader failure. AWS’s September update says the damage affected multiple Availability Zones in ME-SOUTH-1 and exceeded the assumptions behind its regional and multi-AZ services. Replication across zones can improve resilience, but it does not help data that was never copied outside the affected physical footprint.
That is why the practical dividing line is not simply “AWS data in Bahrain” or “AWS data in the UAE.” The decisive question is where a resource was hosted and whether another usable copy existed. AWS’s statement covers resources and data hosted exclusively in the named locations, not every customer resource across both markets.
The recovery path: migration, backups and replacement infrastructure
AWS began recommending migration away from Bahrain after the first Availability Zone was damaged in March. After a second Bahrain zone was disrupted in April, the Region became unavailable. Customers that had remote backups or copies of accessible data could use those assets to re-establish operations in other Regions.
In the UAE, the recovery picture remains mixed: AWS cannot restore exclusively hosted resources in mec1-az2, while replacement work continues in mec1-az1 and mec1-az3. The company has not supplied a concrete restoration date for that work. It said it would provide another UAE update in the coming months.
The next Bahrain milestone is more specific: AWS plans to issue another update in early 2027. That is a communication date, not a promised restoration date.
From the March damage to the September disclosure
AWS began reporting problems in the UAE on March 1. Its chronology later described objects striking mec1-az2, causing sparks and fire. On March 2, AWS said two UAE facilities had been directly struck and that a nearby strike had physically affected a Bahrain facility.
A further disruption affected another Bahrain Availability Zone in April. The September 15 update then set out the lasting operational boundary: resources and data hosted exclusively in Bahrain’s Region could not be restored, while the same applied to exclusively hosted resources and data in UAE zone mec1-az2.
The early incident footage captures the fires, power disruption and connectivity problems reported during the March damage, before AWS published its later recovery assessment.
