Request overload

If many users issue heavy operations at the same time — large searches, exports, AI extractions — request throughput may exceed the system's capacity for a short period. Affected features may respond slowly or temporarily fail until the queue drains.

Hazardous situation: User workflows are halted or delayed because the system received more requests than it could handle in a short period.

• Scalable infrastructure with queue management. Heavy operations are queued and processed asynchronously rather than blocking the user interface.
• Lean searches by design. Smaller, well-scoped searches are more reliable than very large ones. See I have too many poor relevancy results. How can I improve my search? and the search-design guidance in Define a Search Strategy and Understanding Operators.
• Retry guidance. Most short overload events resolve themselves within minutes. See Why does my search execution fail? for retry recommendations.
• Export sizing. Large exports should be sized to your needs — see Export a Search and The export function doesn't work for me.
• Report sustained slowness. If slowness persists beyond what a retry can fix, contact us — see How can I report a problem? and What is the web chat?. For related residual risks see Display delays.