Full refresh vs incremental in dbt - Quick Revision & Key Differences

Recall & Review

beginner

What is a full refresh in dbt?

A full refresh means rebuilding the entire table from scratch every time the model runs. It deletes old data and loads all data fresh.

beginner

What does incremental loading mean in dbt?

Incremental loading means adding only new or changed data to the existing table, instead of rebuilding the whole table.

intermediate

Why use incremental models instead of full refresh?

Incremental models save time and resources by processing only new or updated data, making runs faster and cheaper.

intermediate

When should you use a full refresh in dbt?

Use full refresh when data changes are complex, or when you want to fix data issues by rebuilding the entire table.

advanced

What is a common challenge with incremental models?

Incremental models can be tricky if data updates or deletes happen, because they only add new data and may miss changes.

What happens during a full refresh in dbt?

A Only new data is added to the table

B Data is appended without deleting old data

C The entire table is rebuilt from scratch

D Only data changes are updated

A full refresh deletes the old table and rebuilds it completely with fresh data.

Which is a benefit of incremental models?

A They save time by processing only new data

B Runs are slower but more accurate

C They rebuild the entire table every time

D They delete all old data before loading

Incremental models add only new or changed data, making runs faster and more efficient.

When is a full refresh preferred over incremental?

A When data updates are simple

B When you want to fix data issues by rebuilding

C When you want to save time

D When only new data arrives

Full refresh is used to fix data problems by rebuilding the entire table.

What is a risk of using incremental models?

A They may miss data updates or deletes

B They cannot handle new data

C They delete all data before loading

D They always take longer to run

Incremental models may miss changes if data is updated or deleted because they only add new data.

Which dbt command triggers a full refresh?

A dbt seed

B dbt run --incremental

C dbt test

D dbt run --full-refresh

The --full-refresh flag tells dbt to rebuild tables fully.

Explain the difference between full refresh and incremental models in dbt.

Think about how much data is processed each time.

When would you choose to run a full refresh instead of incremental in your data pipeline?

Consider situations needing a fresh start.