When: 9:24 - 10:22a EDT May 12 (10:24 - 11:22p KST, 1:24 - 2:22p GMT) Impact: Performance degradation for 5-10% of API requests Cause: AWS confirmed that the hardware for a single API instance began to fail shortly after provisioning resulting in slow response times Remediation: Abnormal API instance was removed from the active pool