We are currently using the Oracle Cloud Data Integration pipeline to run tasks consisting of data loader tasks, data integration flow and Oracle SQL procedures. Whenever we restart the pipeline after failure, it starts from the beginning, not from the failed point or state.