connectors

No menu items for this category

Run the MLflow Connector Externally

StagePROD
ML Models

In this section, we provide guides and references to use the MLflow connector.

Configure and schedule MLflow metadata and profiler workflows from the OpenMetadata UI:

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.

OpenMetadata 0.12 or later

To deploy OpenMetadata, check the Deployment guides.

To run the MLflow ingestion, you will need to install:

All connectors are defined as JSON Schemas. Here you can find the structure to create a connection to MLflow.

In order to create and run a Metadata Ingestion workflow, we will follow the steps to create a YAML configuration able to connect to the source, process the Entities if needed, and reach the OpenMetadata server.

The workflow is modeled around the following JSON Schema

This is a sample config for MLflow:

trackingUri: MLflow Experiment tracking URI. E.g., http://localhost:5000

registryUri: MLflow Model registry backend. E.g., mysql+pymysql://mlflow:password@localhost:3307/experiments

The sourceConfig is defined here:

markDeletedMlModels: Set the Mark Deleted Ml Models toggle to flag ml models as soft-deleted if they are not present anymore in the source system.

To send the metadata to OpenMetadata, it needs to be specified as type: metadata-rest.

The main property here is the openMetadataServerConfig, where you can define the host and security provider of your OpenMetadata installation.

Logger Level

You can specify the loggerLevel depending on your needs. If you are trying to troubleshoot an ingestion, running with DEBUG will give you far more traces for identifying issues.

JWT Token

JWT tokens will allow your clients to authenticate against the OpenMetadata server. To enable JWT Tokens, you will get more details here.

You can refer to the JWT Troubleshooting section link for any issues in your JWT configuration.

SSL Configuration

If you have added SSL to the OpenMetadata server, then you will need to handle the certificates when running the ingestion too. You can either set verifySSL to ignore, or have it as validate, which will require you to set the sslConfig.certificatePath with a local path where your ingestion runs that points to the server certificate file.

Find more information on how to troubleshoot SSL issues here.

filename.yaml

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.