Exam Certified Data Engineer Professional topic 1 question 212 discussion

https://docs.databricks.com/en/delta-live-tables/expectations.html "You can maintain data quality rules separately from your pipeline implementations. Databricks recommends storing the rules in a Delta table with each rule categorized by a tag."

Both C and D are correct. Databricks recommends D

Why Option C is Correct DLT native pattern for code reuse — Create a shared notebook with common expectation functions (e.g., check_not_null(), check_valid_date()), then import it into each DLT pipeline notebook using %run or Python imports — exactly how DLT is designed to work

Reasoning: DLT expectations are often repeated across multiple tables (e.g., column not null, value ranges, valid enums). To avoid duplication, the best practice is to factor them out into reusable functions or libraries, and then import them into multiple DLT notebooks. Databricks supports importing shared Python modules or notebooks to centralize and reuse logic.

C, For Sure

The best practice for code reuse is write them once in a shared utility notebook

Reusability of Data Quality Rules: By maintaining the data quality rules in a separate notebook, the team can centralize the logic for expectations and reuse them across multiple tables and pipelines. This approach ensures consistency and reduces duplication of code. Importing as a Library: Databricks allows you to modularize code by creating reusable notebooks or Python files. These can be imported into other notebooks or DLT pipelines, making it easy to apply the same set of expectations across multiple tables.

https://docs.databricks.com/aws/en/dlt/expectation-patterns#portable-and-reusable-expectations Expectations should be stored in the table

D answer is confusing when it comes to use parameter as schema. Having expectations defined as python module and later imported is described in doc: https://docs.databricks.com/aws/en/dlt/expectation-patterns?language=Python%C2%A0Module#portable-and-reusable-expectations

Suretly D

Initially C, currently D

D is what Databricks suggests as of now

To reuse repetitive data quality rules across multiple tables in a Delta Live Tables (DLT) pipeline, the most efficient approach is to maintain these rules in a separate notebook or Python module and import them where needed. This promotes code reusability, maintainability, and consistency