Merge pull request #256 from fbsolo-ms1/update-for-LT

Requested updates to cover CLI-related content . . .

Author: denrea

articles/machine-learning/reference-yaml-feature-store.md

@@ -9,7 +9,7 @@ ms.topic: reference
 author: fbsolo-ms1
 ms.author: franksolomon
 ms.reviewer: qiax
-ms.date: 05/23/2023
+ms.date: 09/12/2024
 ms.custom: cliv2, build-2023
 ---
 
@@ -21,17 +21,19 @@ ms.custom: cliv2, build-2023
 
 ## YAML syntax
 
-
 | Key | Type | Description | Allowed values | Default value |
 |--|--|--|--|--|
 | $schema | string | The YAML schema. If you use the Azure Machine Learning VS Code extension to author the YAML file, including $schema at the top of your file enables you to invoke schema and resource completions. |  |  |
 | name | string | **Required.** Name of the feature store. |  |  |
 | compute_runtime | object | The compute runtime configuration used for materialization job. |  |  |
-| compute_runtime.spark_runtime_version | string | The Azure Machine Learning Spark runtime version. | 3.2 | 3.2 |
+| compute_runtime.spark_runtime_version | string | The Azure Machine Learning Spark runtime version. | 3.4 | 3.4 |
 | offline_store | object |  |  |  |
 | offline_store.type | string | **Required** if offline_store is provided. The type of offline store. Only data lake gen2 type of storage is supported. | azure_data_lake_gen2 |  |
 | offline_store.target | string | **Required** if offline_store is provided. The datalake Gen2 storage URI in the format of `/subscriptions/<subscription_id>/resourceGroups/<resource_group>/providers/Microsoft.Storage/storageAccounts/<account>/blobServices/default/containers/<container>`. |  |  |
-| materialization_identity | object | The user-assigned managed identity that used for the materialization job. This identity needs to be granted necessary roles to access Feature Store service, the data source and the offline storage. |  |  |
+| online_store | object |  |  |  |
+| online_store.type | string | **Required** if online_store is provided. The type of online store. Only redis cache is supported. | redis |  |
+| online_store.target | string | **Required** if online_store is provided. The Redis Cache URI in the format of `/subscriptions/<subscription_id>/resourceGroups/<resource_group>/providers/Microsoft.Cache/Redis/<redis-name>`. |  |  |
+| materialization_identity | object | The user-assigned managed identity that used for the materialization job. This identity needs to be granted necessary roles to access Feature Store service, the data source, and the offline storage. |  |  |
 | materialization_identity.client_id | string | The client ID for your user-assigned managed identity. |  |  |
 | materialization_identity.resource_id | string | The resource ID for your user-assigned managed identity. |  |  |
 | materialization_identity.principal_id | string | the principal ID for your user-assigned managed identity.|  |  |
@@ -48,7 +50,7 @@ You can include other [workspace properties](reference-yaml-workspace.md).
 The `az ml feature-store` command can be used for managing Azure Machine Learning feature store workspaces.
 ## Examples
 
-Examples are available in the [examples GitHub repository](https://github.com/Azure/azureml-examples/tree/main/cli). Some common examples are shown in below.
+Examples are available in the [examples GitHub repository](https://github.com/Azure/azureml-examples/tree/main/cli). Some common examples are shown here:
 
 ## YAML basic
 
@@ -82,6 +84,53 @@ tags:
   foo: bar
 ```
 
+## Configure the online store in the CLI with YAML
+
+```yaml
+$schema: http://azureml/sdk-2-0/FeatureStore.json
+name: mktg-feature-store
+
+compute_runtime:
+  spark_runtime_version: 3.4
+
+online_store:
+  type: redis
+  target: "/subscriptions/<sub-id>/resourceGroups/<rg>/providers/Microsoft.Cache/Redis/<redis-name>"
+
+materialization_identity:
+  client_id: 00001111-aaaa-2222-bbbb-3333cccc4444
+  principal_id: aaaaaaaa-bbbb-cccc-1111-222222222222
+  resource_id: /subscriptions/<sub-id>/resourceGroups/<rg>/providers/Microsoft.ManagedIdentity/userAssignedIdentities/<uai-name>
+
+# Many of workspace parameters will also be supported:
+location: eastus
+display_name: marketing feature store
+tags:
+  foo: bar
+```
+
+## Configure the online store in the CLI with Python
+
+```python
+redis_arm_id = f"/subscriptions/{subscription_id}/resourceGroups/{resource_group_name}/providers/Microsoft.Cache/Redis/{redis_name}"
+online_store = MaterializationStore(type="redis", target=redis_arm_id)
+ 
+fs = FeatureStore(
+    name=featurestore_name,
+    location=location,
+    online_store=online_store,
+)
+ 
+# wait for feature store creation
+fs_poller = ml_client.feature_stores.begin_create(fs)
+
+# move the feature store to a YAML file
+
+yaml_path = root_dir + "/featurestore/featurestore_with_online.yaml"
+fs.dump(yaml_path)
+
+```
+
 ## Next steps
 
 - [Install and use the CLI (v2)](how-to-configure-cli.md)

articles/machine-learning/tutorial-online-materialization-inference.md

@@ -109,6 +109,8 @@ The feature store needs the Azure Cache for Redis as an attached resource, for u
 > - Online store target
 > - Offline store target
 
+For an example that shows how to do this with the SDK, visit the [Tutorial: Different Approaches for Provisioning a Managed Feature Store](https://github.com/Azure/azureml-examples/blob/main/sdk/python/featurestore_sample/notebooks/sdk_and_cli/4.Provision-feature-store.ipynb) resource.
+
 ## Materialize the `accounts` feature set data to online store
 
 ### Enable materialization on the `accounts` feature set
@@ -166,11 +168,11 @@ You can explore the feature materialization status for a feature set from the **
 - Your data can have a maximum of 2,000 *data intervals*. If your data contains more than 2,000 *data intervals*, create a new feature set version.
 - You can provide a list of more than one data statuses (for example, `["None", "Incomplete"]`) in a single backfill job.
 - During backfill, a new materialization job is submitted for each *data interval* that falls in the defined feature window.
-- A new job is not submitted for a *data interval* if a materialization job is already pending, or is running for a *data interval* that hasn't yet been backfilled.
+- A new job isn't submitted for a *data interval* if a materialization job is already pending, or is running for a *data interval* that hasn't yet been backfilled.
 
 ### Updating online materialization store
 - If an online materialization store is to be updated at the feature store level, then all feature sets in the feature store should have online materialization disabled.
-- If online materialization is disabled on a feature set, the materialization status of the already-materialized data in the online materialization store will be reset. This renders the already-materialized data unusable. You must resubmit your materialization jobs after you enable online materialization.
+- If online materialization is disabled on a feature set, the materialization status of the already-materialized data in the online materialization store is reset. This renders the already-materialized data unusable. You must resubmit your materialization jobs after you enable online materialization.
 - If only offline materialization was initially enabled for a feature set, and online materialization is enabled later:
   - The default data materialization status of the data in the online store will be `None`.
   - When the first online materialization job is submitted, the data already materialized in the offline store, if available, is used to calculate online features.