Abstract: Aiming at the problems of high cost,large deployment process,low resource utilization and poor data security of big data processing platform,this paper builds a cloud computing platform for big data processing teaching and scientific research based on OpenStack,which can provide computing,storage services based on virtual machine and Swift container,and provide other func-tions such as personalized customization,automatic deployment,and flexible extension of big data processing platforms like hadoop, spark,storm and others. The platform can effectively utilize existing IT hardware resources,save experimental environment operation and maintenance costs,simplify the deployment and management of big data processing systems,lower the threshold for teaching and research implementation,and provide reliable platform support for big data processing teaching and scientific research experiments.