Looker Validator Implementation Guide

This guide walks you through setting up the Looker Validator tool for your CI/CD process.

Step 1: Set Up a Dedicated Looker User

First, create a dedicated Looker user and API key with the appropriate permissions:

1. Create a permission set:

   • Go to Admin > Roles
   • Create a new permission set named "Validator User"
   • Include these permissions:
     • access_data
     • see_lookml_dashboards
     • see_looks
     • see_user_dashboards
     • develop
     • see_lookml
     • see_sql

2. Create a role:

   • Go to Admin > Roles
   • Create a new role named "Validator User"
   • Add the "Validator User" permission set
   • Select the model set(s) you want to validate

3. Create the user:

   • Go to Admin > Users
   • Create a new user (e.g., "validator@yourcompany.com")
   • Assign the "Validator User" role

4. Generate API credentials:

   • Edit the user
   • Go to API3 Keys
   • Click "New API Key"
   • Save the client ID and client secret

Step 2: Install Looker Validator

You can install the validator in your local environment or in your CI system:

pip install looker-validator

Step 3: Configure the Validator

Create a configuration file looker_validator_config.yaml:

# Looker API credentials
base_url: https://your-looker-instance.cloud.looker.com
client_id: YOUR_CLIENT_ID
client_secret: YOUR_CLIENT_SECRET
project: your_looker_project_name

# Optional global settings
log_dir: logs

# SQL validator settings
concurrency: 10
incremental: true

# Content validator settings
exclude_personal: true
incremental: true

# LookML validator settings
severity: warning

Step 4: Test the Connection

Verify that you can connect to your Looker instance:

looker-validator connect --config-file looker_validator_config.yaml

You should see a success message with your Looker version.

Step 5: Set Up GitHub Action

1. Add secrets to your GitHub repository:

   • Go to your repository settings
   • Click on Secrets > Actions
   • Add the following secrets:
     • LOOKER_BASE_URL: Your Looker instance URL
     • LOOKER_CLIENT_ID: The API client ID
     • LOOKER_CLIENT_SECRET: The API client secret

2. Create workflow file:

   Create a file at .github/workflows/looker-validation.yml:

   name: Looker Validation
   
   on:
     pull_request:
       branches: [ main, master, dev ]
       paths:
         - '**.view.lkml'
         - '**.model.lkml'
         - '**.explore.lkml'
         - '**.dashboard.lookml'
         - '**manifest.lkml'
   
   jobs:
     validate:
       name: Validate Looker Project
       runs-on: ubuntu-latest
       
       steps:
         - name: Checkout code
           uses: actions/checkout@v3
           
         - name: Set up Python
           uses: actions/setup-python@v4
           with:
             python-version: '3.9'
             
         - name: Install looker-validator
           run: |
             python -m pip install --upgrade pip
             pip install looker-validator
             
         - name: Test connection
           run: |
             looker-validator connect \
               --base-url ${{ secrets.LOOKER_BASE_URL }} \
               --client-id ${{ secrets.LOOKER_CLIENT_ID }} \
               --client-secret ${{ secrets.LOOKER_CLIENT_SECRET }} \
               --project ${{ github.event.repository.name }}
         
         - name: Run LookML validation
           run: |
             looker-validator lookml \
               --base-url ${{ secrets.LOOKER_BASE_URL }} \
               --client-id ${{ secrets.LOOKER_CLIENT_ID }} \
               --client-secret ${{ secrets.LOOKER_CLIENT_SECRET }} \
               --project ${{ github.event.repository.name }} \
               --branch ${{ github.head_ref }}
         
         - name: Run SQL validation
           run: |
             looker-validator sql \
               --base-url ${{ secrets.LOOKER_BASE_URL }} \
               --client-id ${{ secrets.LOOKER_CLIENT_ID }} \
               --client-secret ${{ secrets.LOOKER_CLIENT_SECRET }} \
               --project ${{ github.event.repository.name }} \
               --branch ${{ github.head_ref }} \
               --incremental
         
         - name: Run Content validation
           run: |
             looker-validator content \
               --base-url ${{ secrets.LOOKER_BASE_URL }} \
               --client-id ${{ secrets.LOOKER_CLIENT_ID }} \
               --client-secret ${{ secrets.LOOKER_CLIENT_SECRET }} \
               --project ${{ github.event.repository.name }} \
               --branch ${{ github.head_ref }} \
               --incremental \
               --exclude-personal
         
         - name: Run Assert validation
           run: |
             looker-validator assert \
               --base-url ${{ secrets.LOOKER_BASE_URL }} \
               --client-id ${{ secrets.LOOKER_CLIENT_ID }} \
               --client-secret ${{ secrets.LOOKER_CLIENT_SECRET }} \
               --project ${{ github.event.repository.name }} \
               --branch ${{ github.head_ref }}
         
         - name: Upload logs
           if: always()
           uses: actions/upload-artifact@v3
           with:
             name: validation-logs
             path: logs/
   

Step 6: Customize Your Validation Process

Depending on your needs, you may want to customize your validation:

For Large Projects

If you have a large Looker project, consider:

1. Selecting specific models/explores:

   looker-validator sql --explores model_a/* model_b/explore_c
   

2. Increasing concurrency:

   looker-validator sql --concurrency 20
   

3. Running validators in sequence to avoid branch conflicts

For Faster Validations

To speed up validation:

1. Use incremental mode:

   looker-validator sql --incremental
   

2. Use fail-fast for SQL validation:

   looker-validator sql --fail-fast
   

3. Exclude specific explores/folders:

   looker-validator content --explores -model_a/explore_b --folders -33
   

Step 7: Handling Errors

When validation fails, the GitHub Action will fail with a non-zero exit code. The logs will contain detailed information about the failures.

You can view the logs in two ways:

1. GitHub Actions logs: View the output directly in the GitHub Actions interface

2. Downloaded artifacts: Download the logs artifact for detailed error information

Best Practices

1. Ignore dimensions that validly fail:

   dimension: complex_calculation {
     sql: ${TABLE}.value ;;
     tags: ["spectacles: ignore"]
   }
   

2. Use incremental validation to only show new errors

3. Run all validators for comprehensive testing

4. Consider test concurrency in CI environments

5. Set up branch protection rules to prevent merging PRs with validation failures

Troubleshooting

Common Issues

1. Authentication Failures:

   • Verify API credentials
   • Check API user permissions

2. Branch Conflicts:

   • Use separate API users for concurrent validations
   • Add concurrency limits to your GitHub workflow

3. SQL Errors:

   • Check the SQL logs for specific error details
   • Ignore dimensions that validly fail

4. Content Errors:

   • Use --exclude-personal to focus on shared content
   • Check for missing fields referenced by dashboards

Getting Help

If you encounter issues:

1. Run validators with --verbose for detailed logs
2. Check log files in the logs directory
3. Review the GitHub Action output