​
Implementation

An Azure Blob created for every feature. The data type is stored in an index, and the values stored in a keyspace based on their entity. Featureform’s scheduler aims to achieve consistency between Azure’s internal state with the user’s desired state as specified in the metadata service.

​
Configuration

First we have to add a declarative Azure Blob configuration in Python.
azure_blob_config.py
import featureform as ff
azure_blob = ff.register_blob_store(
    name="azure-quickstart",
    description="An azure blob store provider to store offline and inference data" # Optional
    container_name="my_company_container"
    root_path="custom/path/in/container"
    account_name=""
    account_key="" 
)
Once our config file is complete, we can apply it to our Featureform deployment 
featureform apply azure_blob_config.py --host $FEATUREFORM_HOST
We can re-verify that the provider is created by checking the Providers tab of the Feature Registry.

​
Mutable Configuration Fields

  • description
  • account_secret

​
Kubernetes Offline Store

Kubernetes serves as a compute layer for generating training sets, SQL, and Dataframe transformations. To use Kubernetes, a storage layer to store the results of the computation needs to be specified.
azure_blob_config.py
import featureform as ff
k8s_store = ff.register_k8s(
    name="k8s",
    description="Native featureform kubernetes compute",
    store=azure_blob,
    team="featureform-team"
)