Chapter 18 Hooks

At one time, I posed a question to a well-versed data engineer of why the technological space seems to be awash with exotic outlandish names. Although I can’t remember what he replied, but it was after a hearty laughter. Hooks in dbt seems to fit into the calibre of these outlandish names, for there is nothing regarding its purpose in dbt which seems to suggest it will lure and capture your data, read the prey in this case.

You could take hooks as customized SQL models that are out of the box when it comes to dbt. There are various forms of hooks, namely:

  1. pre-hook - executed before a model, snapshot or seed is built.

  2. post-hook - executed after a model, snapshot or seed is built.

  3. on-run-start - executed at the start of implementing the following code executions: dbt build, dbt compile, dbt docs generate, dbt run, dbt seed, dbt snapshot or dbt test.

  4. on-run-end - executed at the end of the following code executions: dbt build, dbt compile, dbt docs generate, dbt run, dbt seed, dbt snapshot or dbt test.

A confession to make: since I have minimal experience using hooks, I shall play in the ‘safe’ zone.

18.1 Post-hooks

The format of writing a hook is:

{{ config(
  post_hook=[
    "<Place your SQl query here>"
  ]
) }}

SELECT * FROM raw_table

We tried creating a simple post-hook but no matter what we tried, dbt kept throwing back errors. Nevertheless, on more on how to set up hooks, see here and here.