Sign up to take part
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
How it different between build upstream only modify and build all upstream. In the lesson it tells that build only modify is different only time, but I need to know that there is any point that are different or not ?
Hi @Parunyu , I believe you are referring to this lesson (https://knowledge.dataiku.com/latest/data-preparation/pipelines/concept-dataset-build-modes.html) and have a question about the difference between the build modes build upstream (only modified) and build all upstream.
Build all upstream is a forced recursive build. No matter whether a dependent dataset is up to date or not, it will recompute all dependencies. Build upstream (only modified), which is the default, will take into account whether a dependent dataset is out of date or not. It won't recompute what doesn't need to be recomputed. So often this option can save computation time.
Hi @Parunyu , I believe you are referring to this lesson (https://knowledge.dataiku.com/latest/data-preparation/pipelines/concept-dataset-build-modes.html) and have a question about the difference between the build modes build upstream (only modified) and build all upstream.
Build all upstream is a forced recursive build. No matter whether a dependent dataset is up to date or not, it will recompute all dependencies. Build upstream (only modified), which is the default, will take into account whether a dependent dataset is out of date or not. It won't recompute what doesn't need to be recomputed. So often this option can save computation time.
@SeanA Understand it! Thanks a lots.