An administrative cleanup command held an "exclusive lock" on key resources, preventing normal application traffic from reading (or writing data in a few cases). This caused a "convoy effect" where incoming API requests queued up behind the lock, resulting in temporary service latency and unavailability.

🛠 Resolution
The issue was resolved by rebooting the database instance to forcibly release the lock. This immediately cleared the pending command and allowed piled-up user requests to process normally. Service has fully recovered and processed queued tasks.

During scheduled maintenance, the primary database experienced a critical Locking Incident.