dbt Analytics Engineering Certification

Question

Consider the following DAG:

model_a model_c model_e model_b model_d model_f

Both model_c and model_d feed into the final layer.

You run:

dbt run --fail-fast

in production with 2 threads. During the run, model_b and model_c are running in parallel when model_b returns an error.

Assume there are no other errors in the model files and that model_c was still running when model_b failed.

Which model or models will successfully build as part of this dbt run? Choose 1 option.

Explanation

The --fail-fast flag stops dbt on the first failure and terminates connections for models still in progress, so model_c is not expected to finish. If model_a already completed, it remains successfully built. The stated DAG does not establish whether model_a completed before the failure, so this result depends on that assumption. (docs.getdbt.com)

Learn more

Failing fast

dbt Analytics Engineering Certification

AstroTutor

Notes

Readiness

QuestionQ31

Community Discussion