Overview
On November 3-4, 2025, customers experienced document-processing delays of up to 30 minutes. This issue affected multiple tenants during peak month-end processing periods was caused by increased processing volumes overwhelming system capacity.
Incident Detail
The issue occurred when document processing volumes increased substantially, causing processing queues to spike significantly higher than regular flow. Investigation revealed that certain cloud services had become unresponsive, preventing automatic scaling mechanisms from handling the increased load effectively. Our engineering teams adjusted service scaling settings. Normal flow resumed at around 10:00 UTC on November 4th, with full resolution confirmed at ~10:30 UTC after all queued documents were processed.
Resolution
The new set of service scaling settings restored normal automatic scaling functionality and cleared the processing backlog. All affected documents were successfully processed once normal operations resumed. We have upgraded monitoring alerts for the processing queues from low to high priority to ensure faster detection of similar issues. Additional improvements include optimizing autoscaling configurations for predictable month-end volume increases and implementing proactive capacity planning to prevent similar incidents during future peak processing periods.