4 ms·
Relevant related post: A data pipeline is just a materialized view: https://nchammas.com/writing/data-pipeline-materialized-view https://nchammas.com/writing/da
by andrenotgiant 6y ago
Relevant related post: A data pipeline is just a materialized view: https://nchammas.com/writing/data-pipeline-materialized-view https://nchammas.com/writing/data-pipeline-materialized-view
- jldlaughlin 6y agoNeat! We're totally on the same page--incremental view maintenance not only makes materialized views a useful building block for data pipelines, it can make them much simpler, too!
- touisteur 6y agoMmmh I've been thinking a lot about generated/computed fields as I wanted to use them in pg. They were introduced in pg12 but only materialized.
- sitkack 6y agoSo it is CTEs all the way down, a bag of dags. A raw table is just a view over the raw data, a cooked view is just a view of views, repeat. Do you use any of the ideas of Noria? Cadence? This is great, but at what COST?!