Configure Power BI lineage mapping

Release version: Australia

Updated March 12, 2026

4 minutes to read

Create a YAML file to map data sources for lineage harvesting.

Before you begin

Role required: admin

About this task

This is an optional task for harvesting lineage information. Create a YAML file and pass it using the Datasource name mapping file option when running the collector.

Set up a YAML file in the following scenarios:

Table 1. Lineage mapping scenarios
Scenario	Details	Action
ODBC Connections in Power BI	You have a data source in Power BI that uses an ODBC connection. In these instances, Power BI doesn’t provide the host or database type of the source	In the YAML file, map the DSN to a specific database host and type. If the database name is missing in the Power BI data source, add the defaultDatabaseName option to the data source in the YAML file
Multiple Server Name Aliases	You have multiple server names (aliases) for the same database instance (host) and the database collector uses a different alias than the one defined in the Power BI connection	Use the YAML file to map the database host to user-specified aliases
Custom SQL Statements	Custom SQL statements are used in Power BI table source definitions. The Power BI collector currently supports connecting to the following database types to resolve lineage from SQL statements: Snowflake, SQL Server, PostgreSQL, Redshift, Oracle, Databricks, Denodo, BigQuery. Lineage resolution for table sources using SQL statements only supports SQL consisting of a single SELECT statement	Configure databases specified in custom SQL statements by including datasourceKey, host, and secure credentials using environment variables

Note:

You can use environment variables in the file for sensitive information such as passwords.

Procedure

Create a YAML file named datasources.yml.

Configure the YAML file based on your scenario.

For multiple server name aliases:

Add the following to map a host alias:

datasources:
 - datasourceKey: "<host or data source key in Power BI source>"
  host: <my-datasource-host>

For example, if your Power BI table source is:

let Source = Snowflake.Database("host-alias.snowflakecomputing.com", "KOS_TEST"),
PowerBiTest_Test_Table = Source{[Schema="POWERBI_TEST",Item="TEST_TABLE"]}[Data]
in PowerBiTest_Test_Table

Then the datasourceKey will be host-alias.snowflakecomputing.com. Your datasources.yml file will look like:

datasources:
- datasourceKey: host-alias.snowflakecomputing.com
host: host-actual.snowflakecomputing.com

For custom SQL statements:

Add the following for databases specified in custom SQL statements. Environment variables are supported:

datasources:
  - datasourceKey: "<host or data source key in Power BI source>"
  OR
 - name: <data source name>
 host: <my-datasource-host>
 databaseUsername: <username> # recommend setting up env variable
 databasePassword: <password> # recommend setting up env variable

Use the following option to specify which databases a datasource configuration applies to. If not provided, it’s assumed that the datasource settings apply to all databases on the given host:

applicableDatabases:
  - <database_name>

For example:

datasources:
- datasourceKey: "example.cpcnqsn422gx.us-east-1.rds.amazonaws.com, 1433"
 host: example.cpcnqsn422gx.us-east-1.rds.amazonaws.com
 databaseUsername: ${DB_USERNAME}
 databasePassword: ${DB_PASSWORD}
 applicableDatabases:
 - 8bank_database

For ODBC connections:

Map the DSN to a specific database host and type. For ODBC connections, list data sources with their corresponding host and database type. If ODBC connections use Odbc.Query, specify the user name and password. If ODBC connections specify the database name, include defaultDatabaseName:
```
datasources:
- name: "Name-for-datasource"
host: <my-datasource-host>
databaseType: <type-of-database>
databaseUsername: <username> # optional
databasePassword: <password> # optional
defaultDatabaseName: <database name> # optional
```
The list of possible databaseTypes are: postgres, redshift, bigquery, oracle, mysql, netezza, snowflake, sqlanywhere, sqlserver, databricks, denodo. Types aren’t case-sensitive but should be a single word with no spaces.

For example:
```
datasources:
- name: "SQL Server DSN Production"
databaseType: sqlserver
host: 8bank-sqlserver.cpetgx.us-east-1.rds.amazonaws.com
```

Optional: Add database-specific configuration options if necessary.

JDBC properties options:

Table 2. JDBC properties
Option	Description
jdbcProperties	Multiple JDBC properties can be specified using a YAML list. The expected format is jdbcProperties: key=value For example: jdbcProperties: encrypt=true For multiple values: `jdbcProperties: - encrypt=true - readOnly=true`

Snowflake database credentials options:

Table 3. Snowflake options
Option	Description
databaseUsername: ${DB_USERNAME}	Required if custom SQL queries are used in Power BI database sources
databasePassword: ${DB_PASSWORD}	Required if SQL queries are used and if a private key isn’t used for authentication to Snowflake
snowflakePrivateKeyFile: privateKeyFile	Required if SQL queries are used and if a private key is used for authentication to Snowflake
snowflakePrivateKeyFilePassword: ${privateKeyFilePassword}	Required if SQL queries are used and if a private key is used for authentication to Snowflake
snowflakeRole: role	Required if SQL queries are used
snowflakeWarehouse: warehouse	Use to override warehouses used in Power BI expressions in the database connection

Databricks database credentials options:

Table 4. Databricks options
Option	Description
databricksHttpPath	Required if the source database is Databricks

Oracle Autonomous Database options:

Option Description

oracleAutonomousDbConnectionString

Table 5. Oracle Autonomous Database options
Option	Description
oracleAutonomousDbConnectionString	Required if the source is an Oracle Autonomous Database. The connection string should be in the format: `jdbc:oracle:thin:<host>:<port>/<service_name>, jdbc:oracle:thin:@(address=(protocol=tcps)(port=<port>)(host:<host>)) (connect_data=(service_name=<service_name>)) (security=(ssl_server_dn_match=yes)), or jdbc:oracle:thin:@alias_name?TNS_ADMIN=/path/to/wallet`

Required if the source is an Oracle Autonomous Database. The connection string should be in the format:

jdbc:oracle:thin:<host>:<port>/<service_name>, 
jdbc:oracle:thin:@(address=(protocol=tcps)(port=<port>)(host:<host>))
(connect_data=(service_name=<service_name>))
(security=(ssl_server_dn_match=yes)), 
or jdbc:oracle:thin:@alias_name?TNS_ADMIN=/path/to/wallet

BigQuery options:

Table 6. BigQuery options
Option	Description
project	Required if the source is a BigQuery table
bigQueryCredentialJsonString	Required for supplying BigQuery credentials

Example for BigQuery database:

datasources:
 - project: "project-name"
databaseType: bigquery
bigQueryCredentialJsonString: '{"key": "value","key": "value"}'

Save the datasources.yml file.
Pass the YAML file to the collector using the --datasource-mapping-file parameter when running the collector.