I am using two dataflow data sources (A1&A2) in a second dataflow (B), where using A1 to filter A2 as part of transformation , it produced prelonged and buggy result in terms of saving the second dataflow and refreshing turnover. However if I change A1 to the original direct datasource (which is sharepoint Online) , it produces much faster and stabler result. Does anyone know if it's an exisiting issue with how dataflow engine works or is there any remedical approach i can adopt?
I prefer using one orignal data source in one dataflow, and using dataflow as the data sources in the following dataflow, which is easier to control source data.
Thanks in advance.