Design and Implementation of a Cloud-Native Platform for Financial Big Data Processing Course.

With the advent of big data, financial technology steps into a more intelligent era. The demand for big data-related financial application development and operation continues to expand. Hadoop and Spark are popular parallel computing frameworks that are adopted as big data processing platforms, but the time cost of their environment deployment is a significant expense for beginners. BDKit, a big data application development and operation platform based on container cloud, is proposed in this paper, aiming to provide customized and scalable cloud service solutions for teaching practice. It is based on Kubernetes ecosystem and relies on the powerful self-repair and load balancing capability provided by Kubernetes to ensure the stable operation of users’ virtual clusters. At the same time, BDKit encapsulates a variety of Docker images to achieve the construction of big data application development and operation environment according to the components required by different users. The services provided by BDKit can greatly reduce the pressure of preparing big data processing environment and help to achieve the purpose of teaching students according to their aptitude.