I have deployed a data pipeline in Google Cloud Data Fusion but it does not work as expected. Is there a way to edit an already deployed data pipeline in Cloud Data Fusion or must it be deleted and rebuilt from scratch and deployed again?
No need to delete and redo, yes any existing pipeline that has been published, can be cloned. This will create an in-memory copy of the pipeline with the same name and opens it within the Cloud Data Fusion Studio.
Go to the Pipeline List of already deployed pipelines select the one you want to "edit" and click the "wheel" and choose duplicate.
At this point, you can rename the pipeline to a unique name and then either save it as a draft or publish it as a new pipeline. As you cannot save over an existing pipeline, all new pipelines need a unique name; a common practice is to increment the names, from test_pipeline01 to test_pipeline02 with each new clone.