DBT Snapshots & Incremental Models Flashcards

What value does dbt store in `dbt_valid_to` for the currently active snapshot row?

Answer: NULL

An open-ended (current) snapshot row has `dbt_valid_to = NULL`, indicating the record is still active.

Which materialization causes dbt to append only new or changed rows instead of rebuilding the whole table?

Answer: incremental

The `incremental` materialization adds or updates only rows that are new since the last run, making large-table processing efficient.

What block in an incremental model filters the source data to only new records?

Answer: {% if is_incremental() %}

`{% if is_incremental() %}` is a dbt Jinja macro that returns true only when the model is running in incremental mode (not full-refresh), allowing you to add a WHERE clause.

What does `dbt run --full-refresh` do to an incremental model?

Answer: Rebuilds the table from scratch, ignoring the existing data

`--full-refresh` drops and recreates an incremental model's table from the ground up, equivalent to the initial build.

Which incremental `strategy` is recommended for Snowflake and BigQuery to efficiently update existing rows?

Answer: merge

The `merge` strategy uses a SQL MERGE statement to upsert rows, updating existing records and inserting new ones in a single pass.

What config key specifies the column(s) used to match existing rows in an incremental merge strategy?

Answer: unique_key

`unique_key` tells dbt which column(s) to use as the join condition when merging new rows into the existing incremental table.