Is making a big data project, only 30 virtual node, data processing 15 TB around every day, because of the limited resources, want to use a Docker + K8s build Hadoop + Spark cluster, whether can meet the requirements on performance?