In this section, we provide guides and references to use the VertexAI connector.
Configure and schedule VertexAI metadata from the OpenMetadata UI:
How to Run the Connector Externally
To run the Ingestion via the UI you'll need to use the OpenMetadata Ingestion Container, which comes shipped with custom Airflow plugins to handle the workflow deployment.
If, instead, you want to manage your workflows externally on your preferred orchestrator, you can check the following docs to run the Ingestion Framework anywhere.
Requirements
Python Requirements
We have support for Python versions 3.9-3.11
To run the VertexAI ingestion, you will need to install:
GCP Permissions
To execute metadata extraction workflow successfully the user or the service account should have enough access to fetch required data. Following table describes the minimum required permissions
| # | GCP Permission | Required For |
|---|---|---|
| 1 | aiplatform.models.get | Metadata Ingestion |
| 2 | aiplatform.models.list | Metadata Ingestion |
Metadata Ingestion
1. Define the YAML Config
This is a sample config for VertexAI:
Source Configuration - Service Connection
credentials: You can authenticate with your vertexai instance using either GCP Credentials Path where you can specify the file path of the service account key or you can pass the values directly by choosing the GCP Credentials Values from the service account key file.
You can checkout this documentation on how to create the service account keys and download it.
gcpConfig:
1. Passing the raw credential values provided by VertexAI. This requires us to provide the following information, all provided by VertexAI:
2. Passing a local file path that contains the credentials:
- gcpCredentialsPath
Location: Location refers to the geographical region where your resources, such as datasets, models, and endpoints, are physically hosted.(e.g. us-central1, europe-west4)
2. Run with the CLI
First, we will need to save the YAML file. Afterward, and with all requirements installed, we can run:
Note that from connector to connector, this recipe will always be the same. By updating the YAML configuration, you will be able to extract metadata from different sources.