Previous incidents

Resolved — the cause was a short-lived database connection spike, now mitigated. All services are operating normally and response times are back to baseline.