Amazon Web Services (AWS): AWS suffers outages as Gulf conflict spills into cloud

Amazon’s Middle‑East infrastructure took a direct hit on March 2, when an “object” struck the mec1‑az2 availability zone in the United Arab Emirates, igniting a blaze that forced local authorities to cut power to the data center, The Register reported. Amazon’s engineers confirmed that the impact generated sparks and fire, and that the facility’s power remained offline for several hours while crews assessed damage to cooling and electrical systems. By 18:46 UTC the outage had spread to a second UAE zone, mec1‑az3, causing “high failure rates for data ingest and egress” across services such as S3 storage, which is normally built to tolerate the loss of a single zone within a region. AWS warned that full restoration could take “at least a day” because repairs require coordination with local authorities and safety checks for on‑site operators.

The disruption was not isolated to the UAE. At 06:56 UTC the same day, AWS’s mes1‑az2 zone in Bahrain reported a “localized power issue” that also left the facility offline, according to the same source. The Bahrain outage coincided with Iranian missile and drone attacks that targeted the US Navy’s Fifth Fleet headquarters in Manama and a high‑rise building believed to be a drone strike, as reported by multiple outlets covering the regional escalation. Amazon’s health dashboard showed that power had not yet been restored to the Bahrain site by 14:23 UTC, and the company reiterated that recovery could take a full day while the damaged infrastructure is repaired.

The cascading failures have already rippled through the SaaS ecosystem in the Gulf. Snowflake, a data‑management platform with a significant customer base in the region, attributed its own service degradation on Sunday to the AWS outage in the UAE, The Register noted. Because many regional enterprises rely on a single AWS region for critical workloads, the loss of two out of three zones in the UAE has forced customers to shift workloads manually or endure prolonged latency, undermining the cloud provider’s usual “single‑zone failure” resilience guarantees.

Analysts observing the incident point to the broader geopolitical risk that now shadows cloud operations in contested regions. Over the past decade, the Middle East has become a strategic hub for big‑tech investment, with Saudi Arabia and the UAE pouring billions into data‑center capacity to diversify away from oil‑dependent economies. The current conflict underscores how quickly physical infrastructure can become a collateral target when state‑level kinetic actions intersect with commercial cloud footprints. Bloomberg’s coverage highlighted that the “objects” that struck the UAE data center were likely linked to the Iranian retaliatory strikes launched after US and Israeli attacks on Tehran earlier in the week.

AWS has not offered a detailed technical post‑mortem, directing inquiries to its public health dashboard, but the incident serves as a stark reminder that cloud reliability is not immune to geopolitical turbulence. As the Gulf remains a flashpoint, enterprises with mission‑critical workloads may need to reconsider multi‑region or multi‑cloud strategies to hedge against similar disruptions in the future.