target_database
Starting in dbt Core v1.9+, this functionality is no longer utilized. Use the database config as an alternative to define a custom database while still respecting the generate_database_name
macro.
Try it now in the dbt Cloud "Latest" release track.
snapshots:
<resource-path>:
+target_database: string
{{ config(
target_database="string"
) }}
Description
The database that dbt should build a snapshot table into.
Notes:
- The specified database must already exist
- On BigQuery, this is analogous to a
project
. - On Redshift, cross-database queries are not possible. If you use this parameter, you will receive the following error. As such, do not use this parameter on Redshift:
Encountered an error:
Runtime Error
Cross-db references not allowed in redshift (raw vs analytics)
Default
By default, dbt will use the target database associated with your profile/connection.
Examples
Build all snapshots in a database named snapshots
snapshots:
+target_database: snapshots
Use a target-aware database
Use the {{ target }}
variable to change which database a snapshot table is built in.
Note: consider whether this use-case is right for you, as downstream refs
will select from the dev
version of a snapshot, which can make it hard to validate models that depend on snapshots.
snapshots:
+target_database: "{% if target.name == 'dev' %}dev{% else %}{{ target.database }}{% endif %}"
Use the same database-naming behavior as models
Leverage the generate_database_name
macro to build snapshots in databases that follow the same naming behavior as your models.
Notes:
- This macro is not available when configuring from the
dbt_project.yml
file, so it must be configured in a snapshot config block. - Consider whether this use-case is right for you, as downstream
refs
will select from thedev
version of a snapshot, which can make it hard to validate models that depend on snapshots.
{{
config(
target_database=generate_database_name('snapshots')
)
}}