aws / me-central-1 / cost and usage report

AWS has declared availability zones in its Bahrain and Dubai regions “hard down” after Iranian strikes damaged data-center infrastructure, and internal communications tell employees to deprioritize those regions. Amazon says both regions remain impaired with limited redundancy and no timeline for normal operations; one zone in each region is hard down while another is impaired but functioning. AWS is helping customers migrate workloads to other regions and reserving capacity for migrations. The attacks, part of wider Iran strikes targeting U.S. tech infrastructure in the Gulf, have also threatened other tech giants and signal risk to cloud resilience and regional continuity for customers operating in the Middle East. This raises broader concerns about geopolitical risk to cloud infrastructure and global service availability.

Amazon reported that availability zones in Bahrain and Dubai went “hard down” after Iranian strikes disrupted regional infrastructure, taking some AWS services offline. The outages affected customers relying on those Middle East zones for compute and networking, highlighting how geopolitical conflict can directly impact cloud availability. The event underscores trade-offs of centralized cloud infrastructure versus dispersed or colocated setups: centralized clouds offer efficiency and shared backbone services but create single points of failure when physical sites are damaged or local staffing is disrupted. For operators and customers, the incident increases focus on multi-region redundancy, geopolitical risk assessment, and disaster recovery planning for mission-critical systems.

AWS announced it will waive all usage charges for March 2026 in its ME-CENTRAL-1 (UAE) region and remove March billing data from Cost and Usage Reports and Cost Explorer. The move follows Iranian drone strikes on March 1 that destroyed two of three availability zones, disrupted about 109 services, and left customers unable to manage resources. Corey Quinn highlights that erasing CUR data removes a key canonical inventory many teams use for compliance, FinOps, and security audits, potentially obscuring what ran in the region during the outage. AWS framed the waiver as automatic and required no customer action; the decision raises operational and auditability concerns even as the physical attack was outside normal AZ design assumptions.

AWS announced it will waive all usage charges for March 2026 in its ME-CENTRAL-1 (UAE) region after Iranian drone strikes destroyed two of three availability zones — and will remove March usage data from Cost and Usage Reports and Cost Explorer. The strikes knocked out control planes and made many services unusable, leaving customers unable to terminate resources; AWS’s email offered no technical explanation and linked only to a brief corporate post. This matters because bills and CUR data are often the canonical inventory and audit trail for cloud resources, relied on by FinOps, security, and compliance teams; erasing that month creates gaps in records and complicates forensic, billing, and compliance reviews. The move raises questions about incident transparency and how cloud providers handle catastrophic, non-standard outages.