You should only update the table in the same pipeline(DLT1) that creates it as that is the pipeline that maintains it. With the assumption that the table data_event is indeed created and maintained in another DLT pipeline, from what you've shared,  you should just move that read of the kafka source into  that pipeline. So everything is in the same DLT pipeline.  In cases where you have to have separate pipelines, you need to trigger 1 after the other (incase of batch mode) or run them continuously (in real time mode), And DLT1 should create a view with a read stream from the event table and finally update the data_event table. Hope this helps.