Coinbase recaps May outage: AWS cascading failures exposed architectural risks

Odaily Odaily reports that Coinbase has released a recap of the massive service outage that occurred on May 7, 2026. The outage lasted approximately 8 hours, with full recovery taking about 12 hours. During this period, trading, deposits, withdrawals, and most core services were unavailable or severely degraded. Coinbase stated that the outage was caused by the simultaneous failure of multiple coolers in the cooling system of a data center in an availability zone (use1-az4) of the AWS us-east-1 region, which triggered a rack thermal protection shutdown, resulting in the EC2 instances and EBS volumes going offline and affecting multiple internet services. During the recovery process, Coinbase's trading matching engine lost its quorum due to the loss of a majority of nodes in its cluster architecture deployed in a single AWS data center. It was necessary to urgently rebuild and restore operation through code adjustments and a new node group, and gradually restart market trading during the recovery process. In addition, a control plane failure occurred in the AWS-managed Kafka (MSK) service, preventing partition master nodes from automatically re-electing. This further disrupted quoting, fee processing, and some settlement and data flow systems, expanding the overall impact. After Coinbase collaborated with the AWS engineering team to perform a manual partition migration, the system gradually returned to normal. Coinbase stated that this incident exposed deficiencies in its cross-availability zone automatic failover capabilities and disaster recovery for managed middleware. The company will upgrade its cross-region hot standby architecture, strengthen regular fault drills, and migrate its Kafka system from a dual-availability zone deployment to a triple-availability zone deployment. It will also work with AWS to advance root cause remediation and improvements.