MATE with dbt 1.4
Changes in MATE with dbt 1.4 public preview
We have introduced a set of changes as part of the dbt 1.4 public preview. Some changes, such as updated versions, are enabled by default, while you will need to enable other changes, such as logical stages, by adding a specific variable.
Switching dbt version
MATE with dbt public preview now supports switching the dbt version using the variable DATAOPS_FEATURE_MATE_DBT_VERSION.
You can now switch between the supported dbt versions 1.4 or 1.0 by setting either value in the variable. You can also install any specific dbt version at runtime.
To know more about the feature, check out Switching dbt version.
Representing logical stages
This new functionality allows organizing nodes in logical stages to better represent the data flow in the generated documentation.
To know more about the feature, check out Logical stages.
Controlling node color
You can also decide on node color in the generated docs along with logical stages using the parameter node_color.
Check out Node colors to know more about the feature.
Support for Python models
This preview also supports Python models, which are introduced in dbt 1.4.
To know more about the feature, check out Using Python to build MATE models.
New TRANSFORM_ACTION: BUILD
We have also added a new TRANSFORM_ACTION: BUILD in MATE with dbt 1.4 preview. The BUILD action will run the RUN, TEST, SNAPSHOT, and EED actions all in a single job.
To know more about the feature, check out Using TRANSFORM_ACTION: BUILD.
Support for YAML selectors
MATE with dbt 1.4 supports the usage of YAML selectors, which allow users to write resource selectors configurations in a human-readable and reusable format.
Check out Resource selectors in YAML to know more about this feature and it's usage.
New parameter TRANSFORM_VARS
MATE with dbt 1.4 supports a new variable TRANSFORM_VARS that allows you to specify key-value pairs as arguments at run-time to dbt.
To know more about this new parameter, check out TRANSFORM_VARS.
Updated version of installed packages
We have also updated the version of bundled packages to the versions listed below:
- Datavault-UK/dbtvault → 0.9.4
- tnightengale/dbt_meta_testing → 0.3.6
- dbt-labs/dbt_utils → 1.0.0
- data-mie/dbt_profiler → 0.6.2
Enabling MATE with dbt 1.4 public preview
You can enable MATE with dbt 1.4 public preview in any branch by adding the following variable to the project variables:
variables:
DATAOPS_RUNNER_IMAGE_TAG: pubpre-dbt1.4
Any subsequent pipeline on the branch will use the orchestrator version, which has the dbt 1.4 preview enabled.