AWS Says It Cannot Restore Some Gulf Data Centre Operations

Amazon Web Services has acknowledged that it cannot restore access to some of its cloud infrastructure in the Gulf after facilities in Bahrain and the United Arab Emirates were severely damaged during the conflict earlier this year. AWS said the damage was extensive enough that some affected resources could no longer be recovered through conventional restoration efforts.

The affected facilities suffered physical damage during strikes in March, disrupting power and connectivity and causing prolonged interruptions to AWS services in the region. In the UAE, one availability zone remains inaccessible, while AWS has also been unable to restore its Bahrain infrastructure to normal operation.

The incident demonstrates the vulnerability of cloud infrastructure to physical events. Cloud computing is often associated with redundancy and resilience, but concentrating critical workloads within a particular geographical region can leave customers exposed when several facilities are affected by the same large-scale disruption.

AWS has been working with affected customers to move workloads to alternative regions. The company has also said that customers with data or applications that were not migrated before the damage could face permanent loss of access to some resources, highlighting the importance of geographically distributed infrastructure for organisations operating critical digital services.

The prolonged disruption is also influencing how governments and technology companies think about the future of data-centre construction in the Gulf. The UAE is reportedly reconsidering the design of its major AI infrastructure projects, including greater geographical distribution and stronger physical protection, as the region seeks to remain a major destination for cloud computing and artificial intelligence investment.