upload-dbfs-temp v0

Overview

Given a file on the local filesystem, this Action uploads the file to a temporary path in DBFS (docs: AWS | Azure | GCP), returns the path of the DBFS tempfile as an Action output, and cleans up the DBFS tempfile at the end of the current GitHub Workflow job.

You can use this Action in combination with databricks/run-notebook to trigger code execution on Databricks for CI (e.g. on pull requests) or CD (e.g. on pushes to master).

Prerequisites

To use this Action, you need a Databricks REST API token to upload your file to DBFS and delete it at the end of workflow job execution. We recommend that you store the token in GitHub Actions secrets to pass it into your GitHub Workflow. The following section lists recommended approaches for token creation by cloud.

AWS

For security reasons, we recommend creating and using a Databricks service principal API token. You can create a service principal, grant the Service Principal token usage permissions, and generate an API token on its behalf.

Azure

For security reasons, we recommend using a Databricks service principal AAD token.

Create an Azure Service Principal

You can:

Install the Azure CLI
Run az login to authenticate with Azure
Run az ad sp create-for-rbac -n <your-service-principal-name> --sdk-auth --scopes /subscriptions/<azure-subscription-id>/resourceGroups/<resource-group-name> --sdk-auth --role contributor, specifying the subscription and resource group of your Azure Databricks workspace, to create a service principal and client secret. Store the resulting JSON output as a GitHub Actions secret named e.g. AZURE_CREDENTIALS
Get the application id of your new service principal by running az ad sp show --id <clientId from previous command output>, using the clientId field from the JSON output of the previous step.
Add your service principal to your workspace. Use the appId output field of the previous step as the applicationId of the service principal in the add-service-principal payload.
Note: The generated Azure token has a default life span of 60 minutes. If you expect your Databricks notebook to take longer than 60 minutes to finish executing, then you must create a token lifetime policy and attach it to your service principal.

Use the Service Principal in your GitHub Workflow

Add the following steps to the start of your GitHub workflow. This will create a new AAD token and save its value in the DATABRICKS_TOKEN environment variable for use in subsequent steps.

- name: Log into Azure
  uses: Azure/login@v1
  with:
    creds: ${{ secrets.AZURE_CREDENTIALS }}
- name: Generate and save AAD token
  id: generate-token
  run: |
    echo "DATABRICKS_TOKEN=$(az account get-access-token \
    --resource=2ff814a6-3304-4ab8-85cb-cd0e6f879c1d \
    --query accessToken -o tsv)" >> $GITHUB_ENV

GCP

For security reasons, we recommend inviting a service user to your Databricks workspace and using their API token. You can invite a service user to your workspace, log into the workspace as the service user, and create a personal access token to pass into your GitHub Workflow.

Usage

See action.yml for the latest interface and docs.

Run a notebook using library dependencies in the current repo

In the workflow below, we build Python code in the current repo into a wheel, use upload-dbfs-temp to upload it to a tempfile in DBFS, then use the databricks/run-notebook Action to run a notebook that depends on the wheel.

name: Upload Python Wheel to DBFS then run notebook using whl.

on:
  pull_request

env:
  DATABRICKS_HOST: https://adb-XXXX.XX.azuredatabricks.net
jobs:
  build:
    runs-on: ubuntu-latest
    steps:
      - name: Checks out the repo
        uses: actions/checkout@v2
      # Obtain an AAD token and use it to upload to Databricks.
      # Note: If running on AWS or GCP, you can directly pass your service principal
      # token via the databricks-host input instead
      - name: Log into Azure
        uses: Azure/login@v1
        with:
          creds: ${{ secrets.AZURE_CREDENTIALS }}
      # Get an AAD token for the service principal,
      # and store it in the DATABRICKS_TOKEN environment variable for use in subsequent steps.
      # We set the `resource` parameter to the programmatic ID for Azure Databricks.
      # See https://docs.microsoft.com/en-us/azure/databricks/dev-tools/api/latest/aad/service-prin-aad-token#--get-an-azure-ad-access-token for details.
      - name: Generate and save AAD token
        id: generate-token
        run: |
          echo "DATABRICKS_TOKEN=$(az account get-access-token \
          --resource=2ff814a6-3304-4ab8-85cb-cd0e6f879c1d \
          --query accessToken -o tsv)" >> $GITHUB_ENV
      - name: Setup python
        uses: actions/setup-python@v2
      - name: Build wheel
        run:
          python setup.py bdist_wheel
      - name: Upload Wheel
        uses: databricks/upload-dbfs-temp@v0
        id: upload_wheel
        with:
          local-path: dist/my-project.whl
      - name: Trigger model training notebook from PR branch
        uses: databricks/run-notebook@v0
        with:
          local-notebook-path: notebooks/deployments/MainNotebook
          # Install the wheel built in the previous step as a library
          # on the cluster used to run our notebook
          libraries-json: >
            [
              { "whl": "${{ steps.upload_wheel.outputs.dbfs-file-path }}" }
            ]
          # The cluster JSON below is for Azure Databricks. On AWS and GCP, set
          # node_type_id to an appropriate node type, e.g. "i3.xlarge" for
          # AWS or "n1-highmem-4" for GCP
          new-cluster-json: >
            {
              "num_workers": 1,
              "spark_version": "10.4.x-scala2.12",
              "node_type_id": "Standard_D3_v2"
            }
          # Grant all users view permission on the notebook results, so that they can
          # see the result of our CI notebook
          access-control-list-json: >
            [
              {
                "group_name": "users",
                "permission_level": "CAN_VIEW"
              }
            ]

License

The scripts and documentation in this project are released under the Apache License, Version 2.0.