User Defined Cloud Function (UDCF) is an external scalar UDF that invokes cloud functions (Function-as-a-Service APIs) directly from VantageCloud Lake. This feature extends the User Defined Function (UDF) interface in the tdextroutine CLI to support associating a cloud function (e.g., AWS Lambda) to an external UDF in the database.
UDCF Considerations
- Currently, only AWS Lambda is supported so the scope of this document is limited to AWS Lambda endpoints.
- Does not support other direct integration with other services such as SageMaker (but can access by Lambda)
- Supported on both primary clusters and compute clusters
- Scalar and Table Operator support for UDCFs
- Lambda must conform to a specific data format
- See UDCF Manifest File for supported manifest.json file property details.
Required Privileges
- The database user must be granted the TD_UDF_ADMIN role to enable/disable the UDCF feature.
- The database user requires CREATE FUNCTION and ALTER FUNCTION privileges on the database to create/replace a UDCF.
- The database user requires EXECUTE FUNCTION privileges on the database to execute a UDCF.
- The database user requires DROP FUNCTION privileges on the database to drop a UDCF.
Function Information
- Set up and enable tdextroutine (XRInstallation) CLI tool
- Establish trust between your AWS Lambda function and VantageCloud Lake environment
- UDCF Manifest File
- Create a UDCF
- List Installed UDCFs
- Get Details about a UDCF
- Replace a UDCF
- Drop a UDCF
- Disable the UDCF Feature
- UDCF data format
- Examples: Working with UDCFs
- UDCF Usage Notes