Skip to main content

Microsoft SQL Server configurations


Ephemeral materialization is not supported due to T-SQL not supporting nested CTEs. It may work in some cases when you're working with very simple ephemeral models.


Tables will, by default, be materialized as a columnstore tables. This requires SQL Server 2017 or newer for on-premise instances or service tier S2 or higher for Azure.

This behaviour can be disabled by setting the as_columnstore configuration option to False.


select *
from ...


By default, dbt-sqlserver will attempt to insert seed files in batches of 400 rows. If this exceeds SQL Server's 2100 parameter limit, the adapter will automatically limit to the highest safe value possible.

To set a different default seed value, you can set the variable max_batch_size in your project configuration.

max_batch_size: 200 # Any integer less than or equal to 2100 will do.


Columns in source tables can not have any constraints. If, for example, any column has a NOT NULL constraint, an error will be thrown.


You can specify indices to be created for your table by specifying post-hooks calling purpose-built macros.

The following macros are available:

  • create_clustered_index(columns, unique=False): columns is a list of columns, unique is an optional boolean (defaults to False).
  • create_nonclustered_index(columns, includes=columns): columns is a list of columns, includes is an optional list of columns to include in the index.
  • drop_all_indexes_on_table(): drops current indices on a table. Only meaningful if the model is incremental.`

Some examples:

"as_columnstore": false,
"materialized": 'table',
"post-hook": [
"{{ create_clustered_index(columns = ['row_id', 'row_id_complement'], unique=True) }}",
"{{ create_nonclustered_index(columns = ['modified_date']) }}",
"{{ create_nonclustered_index(columns = ['row_id'], includes = ['modified_date']) }}",


select *
from ...

Grants with auto provisioning

dbt 1.2 introduced the capability to grant/revoke access using the grants configuration option. In dbt-sqlserver, you can additionally set auto_provision_aad_principals to true in your model configuration if you are using Microsoft Entra ID authentication with an Azure SQL Database or Azure Synapse Dedicated SQL Pool.

This will automatically create the Microsoft Entra ID principal inside your database if it does not exist yet. Note that the principals need to exist in your Microsoft Entra ID, this just makes them available to use in your database.

Principals are not removed again when they are removed from the grants configuration.

auto_provision_aad_principals: true


The following permissions are required for the user executing dbt:

  • CREATE SCHEMA on the database level (or you can create the schema in advance)
  • CREATE TABLE on the database level (or on the user's own schema if the schema is already created)
  • CREATE VIEW on the database level (or on the user's own schema if the schema is already created
  • SELECT on the tables/views being used as dbt sources

The 3 CREATE permissions above are required on the database level if you want to make use of tests or snapshots in dbt. You can work around this by creating the schemas used for testing and snapshots in advance and granting the right roles.

cross-database macros

The following macros are currently not supported:

  • bool_or
  • array_construct
  • array_concat
  • array_append


Many dbt-utils are supported, but require the installation of the tsql_utils dbt package.