Setting up a Pushdown connection

Pushdown processing is currently available for the data sources listed within the tabbed section below. Click a tab to see the Pushdown setup information for an available data source.

  • Databricks
  • Snowflake
  • BigQuery
  • Athena
  • Redshift
  • Trino
  • SQL Server
  • SAP Hana

This section shows you how to set up Pushdown for a Databricks connection.

Requirements

In Collibra DQ

You have:

  • Admin permissions to set up the Databricks connection template and enable Pushdown.
  • Databricks JDBC driver version 2.6.32+ for Standalone deployments of Collibra DQ.

    • Note You do not need to install the Databricks JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In Databricks

You have:

  • A personal access token to establish a connection between Collibra DQ and Databricks.
  • A Databricks workspace admin SQL warehouse that is set up in your Databricks workspace.

    • Note Your SQL warehouse can use either the Hive Metastore or the Delta Lake Unity Catalog for metadata management. Note that when you run Pushdown jobs on tables within the Hive Metastore without adding Unity Catalog to the connection string, Explorer displays all tables available in your Databricks workspace. However, only the tables that are viewable from Databricks under the Hive Metastore can be accessed to run Pushdown jobs on them. We recommend adding a Unity Catalog to your connection URL to run Pushdown jobs on all tables within your Databricks workspace.

  • Read access to the tables on which you plan to run DQ scans.
    • A Databricks workspace admin can enable the Can Read permission in the Permission Settings section of your Databricks workspace.

Configuring Pushdown for Databricks

The following steps show you how to set up the Databricks connection template with Pushdown enabled.

Setting up the Databricks connection template

The following steps show you how to set up the Databricks connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the Databricks tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name String The unique name of your Databricks connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:databricks://<account>.cloud.databricks.com/?db=DATABRICKS_SAMPLE_DATA;ConnCatalog=<catalog-name>

    Important ConnCatalog=<catalog-name> is required if you use Unity Catalog for metadata management. However, it is not required if you use the Hive Metastore.

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 0.

    Yes

    		Driver Name String The driver class name used for your connection, for example, com.databricks.client.jdbc.Driver

    No

    		Source Name String Not applicable.

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown Jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for Databricks connections, go to the official Databricks documentation.

    Yes

    		Pushdown Option

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your Databricks connection.

    Important You must select this option for Pushdown to function.

    No

    		Archive Break Records Option Allows you to write break records directly to a Databricks database or schema.

    Yes

    		Archive Break Records Location String The break records archive location within Databricks. If you do not specify a location, then Collibra DQ uses PUBLIC by default.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver.

    No

    		Driver Properties String

    Optional. The configurable driver properties for your connection.
  3. Click Save.

What's next?

You can now create a Pushdown job on your Databricks connection.

This section shows you how to set up Pushdown for a Snowflake connection.

Running the Snowflake setup script

To run a Snowflake Pushdown job, you need to opt in when setting up your Snowflake connection. Select the Pushdown option in the Connection template to turn on Pushdown capabilities.

Step Details
1 Open a terminal session.
2

Update the following session variables:

Copy 
set dq_username='SERVICE_ACCOUNT_USER';
set dq_password='SERVICE_ACCOUNT_PASSWORD';
set dq_warehouse_name='COLLIBRA_DQ_WH';
set dq_warehouse_size='XSMALL';
set user_database='TARGET_DB';
set dq_role_name='COLLIBRA_DQ_ROLE';

Note dq_password must be uppercase.

Warning Do not update the variables for Collibra DQ.
3

Run as an admin user.

Copy 
USE ROLE ACCOUNTADMIN;
4

Create a user and role for Collibra DQ.

Copy 
CREATE ROLE IF NOT EXISTS identifier($dq_role_name);
CREATE USER IF NOT EXISTS identifier($dq_username) PASSWORD=$dq_password DEFAULT_ROLE=$dq_role_name;
GRANT ROLE identifier($dq_role_name) TO USER identifier($dq_username);
5

Create a virtual warehouse to run Collibra DQ. The virtual warehouse is referenced in the connection URL to direct DQ traffic to it.

Copy 
CREATE WAREHOUSE IF NOT EXISTS identifier($dq_warehouse_name) WAREHOUSE_SIZE=$dq_warehouse_size INITIALLY_SUSPENDED=TRUE
AUTO_SUSPEND = 5 AUTO_RESUME = TRUE;
6

Assign privileges to the Collibra DQ warehouse.

Copy 
GRANT OPERATE, USAGE, MONITOR ON WAREHOUSE identifier($dq_warehouse_name) TO ROLE identifier($dq_role_name);
7

Assign metadata access to your Collibra DQ role.

Copy 
GRANT USAGE,MONITOR on DATABASE identifier($user_database) to identifier($dq_role_name);
GRANT USAGE,MONITOR ON ALL SCHEMAS IN DATABASE identifier($user_database) to identifier($dq_role_name);
8

Update the session variable user_database and grant read access to objects in the user database. Run this portion for each target database within which you want to run DQ checks.

Copy 
USE DATABASE identifier($user_database);
GRANT SELECT ON ALL TABLES IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
GRANT SELECT ON ALL VIEWS IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
GRANT SELECT ON ALL EXTERNAL TABLES IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
GRANT SELECT ON ALL STREAMS IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
                                ​
GRANT SELECT ON FUTURE TABLES IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
GRANT SELECT ON FUTURE VIEWS IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
GRANT SELECT ON FUTURE EXTERNAL TABLES IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);
GRANT SELECT ON FUTURE STREAMS IN DATABASE identifier($user_database) TO ROLE identifier($dq_role_name);

Warning Please ensure the SQL variables are updated in the above script before proceeding.

Setting up the Snowflake Connection template

The following steps show you how to set up the Snowflake connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the Snowflake tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name Text The unique name of your Snowflake connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:snowflake://<account>.snowflakecomputing.com/?db=SNOWFLAKE_SAMPLE_DATA&warehouse=COLLIBRA_DQ_WH

    Yes

    		Port Integer The port number to establish a connection to the datasource.

    Yes

    		Driver Name Text The driver class name used for your connection, for example, com.snowflake.client.jdbc.SnowflakeDriver

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    You do not need to select an option because Pushdown Jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. See the Snowflake documentation for more information about authentication methods for Snowflake connections.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your Snowflake connection.

    Warning You must select this option for Pushdown to function.

    No

    		Archive Break Records Option

    Allows you to write break records directly to a Snowflake database or schema.

    Important Ensure that you have the necessary privileges to write break records to Snowflake.

    Yes

    		Driver Location File path A file path which designates the name and location of the driver.

    Yes

    		Driver Properties String

    The configurable driver properties for your connection. Add the following driver property:

    CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX=TRUE

    The following image shows an example of a Snowflake connection with Pushdown turned on and additional Driver Properties.

    screenshot of a sample snowflake pushdown connection with "pushdown" and "driver properties" highlighted

  3. Click Save.

Known limitations

  • There is a limitation where Snowflake Pushdown jobs with queries containing REGEXP_REPLACE fail because Snowflake does not currently support arguments with collation specifications. There are two known workarounds:
    • Make a clone of the table that is synced in Snowflake and has no collation set.
    • Use Pullup mode instead of Pushdown to avoid this Snowflake limitation.

What's next?

You can now create a Pushdown job on your Snowflake connection.

This section shows you how to set up Pushdown for a BigQuery connection.

Requirements

In Collibra DQ

  • You have admin permissions (ROLE_ADMIN) to set up the BigQuery connection template and enable Pushdown.
  • You have BigQuery JDBC driver version 1.3.2.1003 for Standalone deployments of Collibra DQ.

    • Note You do not need to install the BigQuery JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In BigQuery

  • You have Viewer (roles/viewer) permissions.
  • You have Reader access to BigQuery dataset tables.
  • You have Writer access to materialized view datasets.

Configuring Pushdown for BigQuery

The following steps show you how to set up the BigQuery connection template with Pushdown enabled.

Setting up the BigQuery connection template

The following steps show you how to set up the BigQuery connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the BigQuery tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name String The unique name of your BigQuery connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;OAuthType=0;ProjectId=owl-hadoop-cdh;OAuthServiceAcctEmail=<service-account-email>;TimeOut=3600

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 443.

    Yes

    		Driver Name String The driver class name used for your connection, for example, com.simba.googlebigquery.jdbc42.Driver

    Note The cdata.jdbc.googlebigquery.GoogleBigQueryDriver does not support Pushdown.

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for BigQuery connections, go to the official BigQuery documentation.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your BigQuery connection.

    Important You must select this option for Pushdown to function.

    No

    		Archive Break Records Option

    Allows you to write break records directly to a BigQuery database or schema.

    Note As of Collibra DQ version 2023.08, only rule break records are eligible for source archival.

    Yes

    		Archive Break Records Location String

    The break records archive location within BigQuery. If you do not specify a location, then Collibra DQ uses PUBLIC by default.

    Note Archive Break Records Location is only available and required when the Archive Break Records option is selected.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver. For example, /opt/owl/drivers/bigquery

    No

    		Driver Properties String

    Optional. The configurable driver properties of your connection.

  3. Click Save.

Limitations

Rule validation for BigQuery Pushdown connections is not supported. However, you can manually execute SQL queries from the BigQuery console on GCP as a possible workaround.

What's next?

You can now create a Pushdown job on your BigQuery connection.

This section shows you how to set up Pushdown for a Amazon Athena connection.

Requirements

In Collibra DQ

  • You have admin permissions (ROLE_ADMIN) to set up the Athena connection template and enable Pushdown.
  • You have Athena JDBC driver version 2.0.35.1000 for Standalone deployments of Collibra DQ.

    • Note You do not need to install the Athena JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In Athena

  • You have Read access to your Glue Catalog and S3 query result bucket location.
  • You have Write access to your S3 query result bucket location.
  • You have Read access to the Athena workgroup.

Configuring Pushdown for Athena

The following steps show you how to set up the Athena connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the Athena tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name String The unique name of your Athena connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:awsathena://AwsRegion=${region};S3OutputLocation=s3://{output_location};MetadataRetrievalMethod=Query

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 444.

    Yes

    		Driver Name String The driver class name of your connection, for example, com.simba.athena.jdbc.Driver

    Note The cdata.jdbc.amazonathena driver class does not support Pushdown.

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for Athena connections, go to the official Athena documentation.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your Athena connection.

    Important You must select this option for Pushdown to function.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver. For example, /opt/owl/drivers/athena/

    No

    		Driver Properties String

    Optional. The configurable driver properties of your connection.

  3. Click Save.

What's next?

You can now create a Pushdown job on your Athena connection.

Known limitations

  • As of the Collibra DQ 2023.09, only AdaptiveRules are available from the Add Layers step.
  • Archive Break Records is not currently supported.

This section shows you how to set up Pushdown for a Redshift connection.

Requirements

In Collibra DQ

  • You have admin permissions (ROLE_ADMIN) to set up the Redshift connection template and enable Pushdown.
  • You have Redshift JDBC driver version 2.1.0.9 for Standalone deployments of Collibra DQ.

    • Note You do not need to install the Redshift JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In Redshift

  • You have READ access on your Redshift database tables.
  • You have USAGE access on your Redshift schema.

Configuring Pushdown for Redshift

The following steps show you how to set up the Redshift connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the Redshift tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name String The unique name of your Redshift connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:redshift://<host>:5439/dev

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 5439.

    Yes

    		Driver Name String The driver class name used for your connection, for example, com.amazon.redshift.jdbc42.Driver

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for Redshift connections, go to the official Redshift documentation.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your Redshift connection.

    Important You must select this option for Pushdown to function.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver. For example, /opt/owl/drivers/redshift/

    No

    		Driver Properties String

    Optional. The configurable driver properties of your connection.

  3. Click Save.

What's next?

You can now create a Pushdown job on your Redshift connection.

Known limitations

  • As of the Collibra DQ 2023.09, only AdaptiveRules are available from the Add Layers step.

This section shows you how to set up Pushdown for a Trino connection.

Requirements

In Collibra DQ

  • You have admin permissions (ROLE_ADMIN) to set up the Trino connection template and enable Pushdown.
  • You have Trino JDBC driver version 405.0 for Standalone deployments of Collibra DQ.

    • Note You do not need to install the Trino JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In Trino

  • You have Read access on your Trino database tables.

Configuring Pushdown for Trino

The following steps show you how to set up the Trino connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the Trino tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name String The unique name of your Trino connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:trino://${host}:443/<catalog>/<schema>&source=jdbc:presto

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 8080.

    Yes

    		Driver Name String The driver class name used for your connection, for example, io.trino.jdbc.TrinoDriver

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for Trino connections, go to the official Trino documentation.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your Trino connection.

    Important You must select this option for Pushdown to function.

    No

    		Archive Break Records Option

    Allows you to write break records directly to a Trino database or schema.

    Note As of Collibra DQ version 2023.10, only rule break records are eligible for source archival.

    Yes

    		Archive Break Records Location String

    The break records archive location within Trino. If you do not specify a location, then Collibra DQ uses PUBLIC by default.

    Note Archive Break Records Location is only available and required when the Archive Break Records option is selected.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver. For example, /opt/owl/drivers/trino

    No

    		Driver Properties String

    Optional. The configurable driver properties of your connection.

  3. Click Save.

Turning on Pushdown for Trino

To ensure that Pushdown for Trino is appropriately configured, verify that PUSHDOWN_JOB_ENABLED is set to TRUE on the Application Config page of the Admin Console.

  1. On the main menu, click , and then click Admin Console.
  2. Click Application Config under the Configuration Settings section.

    3. Tip If PUSHDOWN_JOB_ENABLED is already set to TRUE, you can skip this section.

  3. If PUSHDOWN_JOB_ENABLED is set to FALSE, click the to the right of the Value column.
  4. In the Edit Configuration modal, enter TRUE into the Value input field.
  5. Click Save.

What's next?

You can now create a Pushdown job on your Trino connection.

This section shows you how to set up Pushdown for a SQL Server connection.

Requirements

In Collibra DQ

  • You have admin permissions (ROLE_ADMIN) to set up the SQL Server connection template and enable Pushdown.
  • You have the SQL Server JDBC driver for Standalone deployments of Collibra DQ.

    • Note You do not need to install the SQL Server JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In SQL Server

  • You have Read access on your SQL Server database tables.

Configuring Pushdown for SQL Server

The following steps show you how to set up the SQL Server connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click the SQL Server tile, then enter the required information.
    3. Required Property Type Description

    Yes

    		Name String The unique name of your SQL Server connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:sqlserver://${host}:1433/databaseName=${dbName};encrypt=true;trustServerCertificate=true;

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 1433.

    Yes

    		Driver Name String The driver class name used for your connection, for example, com.microsoft.sqlserver.jdbc.SQLServerDriver

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for SQL Server connections, go to the official SQL Server documentation.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your SQL Server connection.

    Important You must select this option for Pushdown to function.

    No

    		Archive Break Records Option

    Allows you to write break records directly to a SQL Server database or schema.

    Note As of Collibra DQ version 2024.02, only rule break records are eligible for source archival.

    Yes

    		Archive Break Records Location String

    The break records archive location within SQL Server. If you do not specify a location, then Collibra DQ uses SQLS by default.

    Note Archive Break Records Location is only available and required when the Archive Break Records option is selected.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver. For example, /opt/owl/drivers/mssql/

    No

    		Driver Properties String

    Optional. The configurable driver properties of your connection.

  3. Click Save.

What's next?

You can now create a Pushdown job on your SQL Server connection.

Known limitations

  • As of the Collibra DQ 2024.02 release, only AdaptiveRules are available from the Add Layers step.
  • Only native Freeform SQL (PQ) is supported for SQLF rules, meaning that queries in the format of @dataset (TLQ) or @job-name (LQ) are not supported. Only queries that use native (PQ) are supported.
  • The use of the 'TOP(limit)' clause for the source is not supported when Link ID is enabled for SQLF rules. This is because the 'TOP N' specification should be placed at the beginning of the query. When Link ID is enabled for rules, it is added to the beginning of the query, causing the Insert query to fail.

This section shows you how to set up Pushdown for a SAP HANA connection.

Requirements

In Collibra DQ

  • You have admin permissions (ROLE_ADMIN) to set up the SAP HANA connection template and enable Pushdown.
  • You have the SAP HANA JDBC driver for Standalone deployments of Collibra DQ.

    • Note You do not need to install the SAP HANA JDBC driver for Cloud native deployments of Collibra DQ because it is already included in the installation package.

In SAP HANA

  • You have Read access on the SAP HANA database tables on which you run data quality.

Configuring Pushdown for SAP HANA

The following steps show you how to set up the SAP HANA connection template with Pushdown enabled.

  1. On the main menu, click , and then click Connections.
  2. Click Add Connection, then click SAP HANA.
  3. Enter the required information.
    4. Required Property Type Description

    Yes

    		Name String The unique name of your SAP HANA connection.

    Yes

    		Connection URL String The base connection string, for example, jdbc:sap://${host}:${port}/

    Yes

    		Port Integer The port number to establish a connection to the datasource. The default port is 39015.

    Yes

    		Driver Name String The driver class name used for your connection, for example, com.sap.db.jdbc.Driver

    No

    		Source Name String N/A

    No

    		Target Agent Option

    The Agent used to submit your DQ Job.

    Note You do not need to select an option because Pushdown jobs do not require an Agent to submit them to the data warehouse.

    Yes

    		Auth Type Option The method to authenticate your connection. For more information on authentication methods for SAP HANA connections, go to the official SAP HANAdocumentation.

    Yes

    		Pushdown Checkbox

    Switches Pushdown processing on or off.

    Select the checkbox option to turn on Pushdown for your SAP HANA connection.

    Important You must select this option for Pushdown to function.

    No

    		Archive Break Records Option

    Allows you to write break records directly to a SAP HANA database or schema.

    Note As of Collibra DQ version 2024.05, only rule break records are eligible for source archival.

    Yes

    		Archive Break Records Location String

    The break records archive location within SAP HANA. If you do not specify a location, then Collibra DQ uses SAPHANADB by default.

    Your default output schema may be one of the following:

    • SAP<SID>
    • SAPSR3
    • SAPBAP<#>
    • SAPHANADB

    Note Archive Break Records Location is only available and required when the Archive Break Records option is selected.

    Yes

    		Driver Location File path A file path that designates the name and location of the driver. For example, /opt/owl/drivers/sap/

    No

    		Driver Properties String

    Optional. The configurable driver properties of your connection.

  4. Click Save.

The following table contains an overview of the data quality features currently available for Pushdown for SAP HANA.

What's next?

You can now create a Pushdown job on your SAP HANA connection.