the current cloud storage platform architecture is developed on the basis of individual PC environment, with poor interactivity, low utilization rate and large memory in large software systems. To this end, a multi - tenant Intelligent Cloud Storage Platform Architecture Based on micro - application is proposed. A multi tenant Intelligent Cloud Storage Platform Architecture Based on micro application is designed. The realization of data management, in accordance with the characteristics of hot and cold data, distinguishing method of large scale data were distinguished by the hot and cold data, according to the tenant behavior, calculate the value of data per unit time, time value of the data are estimated by the exponential smoothing method, improve the accuracy of data access. After the cold and hot data are distinguished, the multi tenant monitoring data is migrated to the current platform architecture by using the Gossip method to ensure the service. Micro applications are developed on the UAP platform, and UAP microapplication development platform is applied to effectively integrate micro application related components, and accomplish the automatic cluster planning of large-scale micro applications. The experimental results show that the proposed method has high memory usage and efficiency because of the application of micro - application.