Service Incident Postmortem

Incident date: August 6, 2026
Incident window: Approximately 1:05–1:20 AM PDT

Summary

At approximately 1:05 AM PDT, the production database server was inadvertently shut down while maintenance was being performed on a separate server.

Because our services depend on the database platform, customers experienced a service interruption or critically degraded functionality while the database server was offline and SQL Server was unavailable.

Our team investigated the interruption, identified that the database server had been shut down, and initiated recovery. The server completed its startup process at approximately 1:20 AM PDT, after which database connectivity and dependent services were restored.

Timeline

• Approximately 1:05 AM PDT: The production database server was inadvertently shut down.
• 1:05–1:20 AM PDT: Our team investigated the service interruption, identified the cause, and initiated recovery of the database server.
• Approximately 1:20 AM PDT: The database server completed startup, SQL Server became available, and dependent application services recovered.

Root Cause

The incident was caused by human error during maintenance activity involving another server. An action intended for the maintenance target was mistakenly applied to the production database server.

Corrective Actions

We are reviewing our maintenance procedures and administrative safeguards to reduce the likelihood of a similar error. Corrective actions will include strengthening server-identification and verification steps before disruptive actions are performed, along with reviewing available Azure controls that can provide additional protection for critical production resources.

We apologize for the disruption and appreciate your patience while service was restored.