On the evening of May 7, 2026, AWS reported problems in one Availability Zone, use1-az4, of its US-EAST-1 region in Northern Virginia. AWS said EC2 instances and EBS storage volumes lost power during a "thermal event" and that dependent services in the zone could also be affected. The Register reported that AWS brought additional cooling online and began recovering racks by about 10 p.m. Pacific, after warning that restoration was slower than expected. AWS shifted traffic away from the zone and advised customers to move workloads to other zones in the region.
The impact was limited to one zone, yet StatusGator counted more than 150 downstream cloud services affected, including Coinbase, Datadog, MongoDB, Supabase, HubSpot and Canva. Many business applications you buy run on AWS behind the scenes, so an AWS problem can surface as an outage in a tool your team uses daily.
If you run workloads on AWS, confirm that production systems span at least two Availability Zones and that a single-zone loss has actually been tested. For SaaS vendors you depend on, ask which cloud and region they run in, whether they are multi-zone or multi-region, and what their contract says about uptime and credits.
US-EAST-1 also had major outages in 2021 and October 2025. AWS executives have said the region is not inherently more fragile but runs at larger scale. For critical systems, that history is worth weighing when choosing a primary region.
Sources
Does this affect your business?
Our advisors can help you figure out the right move, at no cost.