Manage the environment through public APIs

The Microsoft Fabric REST API provides a service endpoint for the create, read, update, and delete (CRUD) operations of a Fabric item. This article describes the available environment REST APIs and their usage.

Summary of available environment APIs

Learn more about the environment public APIs in Item APIs - Environment

Environment public API use cases

This section walks you through several common scenarios when dealing with environment. You can replace the {WORKSPACE_ID} and {ARTIFACT_ID} in the following examples with appropriate values.

Create a new environment

You can create a new empty environment using the following API.

• Sample request

  POST https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments
  
  {
      "displayName": "Environment_1",
      "description": "An environment description"
  }
  

Manage staging libraries

You can use the upload/delete staging libraries APIs to manage the library section in the environment

Check the published libraries for the environment

Before adding/deleting library, you can use the get published libraries API to check what libraries are currently effective.

• Sample request

  GET https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/libraries
  

• Sample response

  {
      "customLibraries": {
          "wheelFiles": [
              "samplewheel-0.18.0-py2.py3-none-any.whl"
          ],
          "pyFiles": [
              "samplepython.py"
          ],
          "jarFiles": [
              "samplejar.jar"
          ],
          "rTarFiles": [
              "sampleR.tar.gz"
          ]
      },
      "environmentYml": "dependencies:\r\n- pip:\r\n  - matplotlib==3.4.3"
  }
  

Uploading the libraries

The API for uploading staging library accepts one file at a time. The supported file types are .whl, .jar, .tar.gz, .py and environment.yml for public library. You can specify the file via the multipart/form-data content-type.

• Sample requests

  POST https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/libraries
  

Deleting the libraries

By specifying the full library file name with the type suffix, you can delete one library at a time.

• Sample requests

  DELETE https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/libraries?libraryToDelete=fuzzywuzzy-0.18.0-py2.py3-none-any.whl
  

Manage staging Spark compute

You can use the update staging Spark compute to manage the Spark compute.

Check the published Spark compute for the environment

Before changing the configurations for the environment, you can use the get published Spark compute API to check what Spark compute configurations are currently effective.

• Sample request

  GET https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/sparkcompute
  

• Sample response

  {
      "instancePool": {
          "name": "Starter Pool",
          "type": "Workspace"
      },
      "driverCores": 4,
      "driverMemory": "56g",
      "executorCores": 4,
      "executorMemory": "56g",
      "dynamicExecutorAllocation": {
          "enabled": false,
          "minExecutors": 1,
          "maxExecutors": 1
      },
      "sparkProperties": {
          "spark.acls.enable": "false"
      },
      "runtimeVersion": "1.2"
  }
  

Update the compute configurations

You can update the Spark runtime, switch to another pool, refine compute configuration and add/remove Spark properties through editing the request body of this API.

If you want to remove an existing Spark property, you need to specify the value as null with the key that you want to remove, as showing in the following example.

• Sample request

  PATCH https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/sparkcompute
  
  {
      "instancePool": {
          "name": "Starter Pool",
          "type": "Workspace"
      },
      "driverCores": 4,
      "driverMemory": "56g",
      "executorCores": 4,
      "executorMemory": "56g",
      "dynamicExecutorAllocation": {
          "enabled": false,
          "minExecutors": 1,
          "maxExecutors": 1
      },
      "sparkProperties": {
          "spark.acls.enable": null
      },
      "runtimeVersion": "1.2"
  }
  

Make the changes effective

Using the following sets of APIs to publish the changes.

Prepare for a publish

The environment can accept one publish at a time. Before publishing your environment, you can validate the status of the environment and have a final review of the staging changes. Once the environment is published successfully, all configurations in the staging state become effective.

• Step 1: get the metadata of the environment

  GET https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/
  

  In the response body, you can tell clearly the state of the environment. Make sure there is no ongoing publish before you move to next step.

• Step 2: get the staging libraries/Spark compute to have a final review

  GET https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/libraries
  
  GET https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/sparkcompute
  

Trigger the publish of the environment

The changes you made in for the staging libraries and Spark compute are cached but require publishing to become effective. Follow the following example to trigger the publish.

• Sample request

  POST https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/publish
  

• Sample response

  {
      "publishDetails":
      {
          "state": "Running",
          "targetVersion": "46838a80-5450-4414-bea0-40fb6f3e0c0d",
          "startTime": "2024-03-29T14:17:09.0697022Z",
          "componentPublishInfo": {
              "sparkLibraries": {
                  "state": "Running"
              },
              "sparkSettings": {
                  "state": "Running"
              }
          }
      }
  }   
  

During the publish, you can also call following API to cancel it.

• Sample request

  POST https://api.fabric.microsoft.com/v1/workspaces/{{WORKSPACE_ID}}/environments/{{ARTIFACT_ID}}/staging/cancelPublish
  

Related content

• Create, configure, and use an environment in Microsoft Fabric.
• Item APIs - Environment.