Connect a data source¶
CatalystData currently supports four data sources. Each one connects via ADBC (Arrow Database Connectivity) — there are no drivers for you to install; that's handled on our side.
More connectors are coming
SQL Server, Microsoft Fabric, Snowflake, and Databricks are supported today. More connectors are on the roadmap — talk to your CatalystData contact if a data source you need isn't listed here yet.
Overview¶
| Connector | connector_type |
Default port | Auth types |
|---|---|---|---|
| SQL Server | sqlserver |
1433 | sql_auth; Entra ID: azure_ad_service_principal, azure_ad_managed_identity, azure_ad_default, azure_ad_access_token |
| Microsoft Fabric | fabric |
1433 | Entra ID only: azure_ad_service_principal, azure_ad_managed_identity, azure_ad_default, azure_ad_access_token |
| Snowflake | snowflake |
443 | password, key_pair |
| Databricks | databricks |
443 | token, oauth |
SQL Server (sqlserver)¶
| Auth type | Description | Fields you will enter |
|---|---|---|
sql_auth |
SQL Server username and password | Username, Password |
azure_ad_service_principal |
Entra ID app registration with a client secret or certificate | Username (client ID), Password (secret) |
azure_ad_managed_identity |
System- or user-assigned managed identity | Username (client ID) — leave blank for a system-assigned identity |
azure_ad_default |
The ambient Entra ID credential chain | Nothing extra |
azure_ad_access_token |
CatalystData acquires a token itself and forwards it as the password | Nothing extra |
Extra parameters available on this connector: application_client_id and
client_cert_path (for azure_ad_service_principal), client_id (for
azure_ad_managed_identity, an alternative to the Username field),
encrypt (true/false/strict), TrustServerCertificate, connection
timeout, and instance (a named instance).
Microsoft Fabric (fabric)¶
Connects to a Fabric Warehouse or Lakehouse SQL analytics endpoint over the same TDS protocol as SQL Server, on port 1433. Entra ID authentication only — Fabric does not support SQL authentication.
- Host: the SQL connection string host from the Fabric item's Settings →
SQL endpoint page, for example
workspace.datawarehouse.fabric.microsoft.com - Database: the Warehouse or Lakehouse name (its "Initial Catalog")
- Prerequisites: the Fabric tenant setting "Service principals can use Fabric APIs" must be enabled, and the service principal needs Contributor (or an equivalent role) on the workspace or the specific item
Auth types and extra parameters are the same as SQL Server's Entra ID options above.
Snowflake (snowflake)¶
| Auth type | Fields you will enter |
|---|---|
password |
Username, Password |
key_pair |
Username — private key details go in extra parameters |
Extra parameters: schema, warehouse, role, login_timeout (all auth
types), plus private_key_file and private_key_passphrase for key_pair.
Databricks (databricks)¶
Set Host to the workspace hostname, for example
adb-1234567890.1.azuredatabricks.net.
| Auth type | Fields you will enter |
|---|---|
token |
Password (a personal access token) |
oauth |
Username (client ID), Password (client secret) — OAuth machine-to-machine |
Required extra parameter: http_path — the SQL warehouse or cluster's
HTTP path. Optional: schema (default catalog/schema).