Dynamic tables
Snowflake dynamic tables integrate transformational logic with data pipelines into a single, self-orchestrating object. Dynamic tables behave similarly to materialized views but have much fewer limitations in their transformational logic and give users control over the latency of data refreshes. They are an alternative to external ETL/ELT processes requiring third-party tools to manage data movement and internally managed Snowflake tasks because dynamic tables encapsulate the transformation logic and data refresh behavior within a single object.
A dynamic table consists of a SQL transformation (like a view), a specified warehouse, and a target lag parameter, which defines the maximum threshold for downstream data freshness relative to upstream data changes. Snowflake automatically ensures that the table stays updated by incrementally processing changes in source data. This built-in automation reduces the complexity of managing data pipelines, as the services layer...
