As the IoT data type and data volume increase

a new challenge is proposed to the storage and calculation of IoT data.To meet the storage and calculation demands of the massive unstructured IoT data

a scheme that NoSQL database technology combined with MapReduce programming framework has been presented in this paper. Using the MongoDB

a typical database of NoSQL

as the main database to store massive unstructured IoT data. At the same time

the Hadoop MapReduce was adopted as the calculation framework for IoT data analysis and processing. Through the overlapping deployment of MongoDB cluster and Hadoop cluster

the cost of data transmission is reduced

and a set of high availability and high performance large data processing platform is established. The analysis of the experiment that using this scheme to deal with massive IoT data has verified the high efficiency and high efficiency of this method.