The large cloud and network monitoring system for enterprise customers provides users with services such as cloud and network monitoring as well as application performance monitoring in a multi-cloud/multi-network environment and provides enterprises with integrated monitoring and maintenance management. It has high requirements for monitoring sensitivity. The factors affecting the system exist in many links such as the data collection, data processing, data storage, data presentation, data cache, and message queue. This study analyses the architecture of the cloud and network monitoring system layer by layer and improves the overall monitoring sensitivity of the system through data analysis and corresponding solution design.