Microsoft Docs

Deploy Azure Analysis Services

This section covers the various methods and considerations for deploying Azure Analysis Services (AAS) models to your Azure environment. Effective deployment ensures your analytical solutions are readily available to users and integrate seamlessly with your data infrastructure.

Deployment Methods

You can deploy your Analysis Services models using several approaches, each suited for different scenarios and workflows:

• SQL Server Data Tools (SSDT): SSDT is an extension for Visual Studio that provides a rich development environment for creating, deploying, and managing Analysis Services models. It's the primary tool for developing tabular and multidimensional models.

  Steps:

  1. Develop your model in Visual Studio with SSDT installed.
  2. Build the project to create the model assemblies.
  3. Right-click the project in Solution Explorer and select "Deploy".
  4. Configure the deployment target server and database name.
• Azure Portal: While you primarily develop models offline, the Azure portal offers capabilities to manage and provision your Analysis Services instance. You can also deploy basic models or synchronize data using other tools.
• Azure PowerShell and Azure CLI: For automation and scripting, you can use Azure PowerShell cmdlets or Azure CLI commands to deploy models programmatically. This is ideal for CI/CD pipelines.

  Example using Azure PowerShell:

  
  # Connect to Azure
  Connect-AzAccount
  
  # Set your context
  Set-AzContext -SubscriptionId "YOUR_SUBSCRIPTION_ID"
  
  # Deploy Analysis Services model (requires Azure.AnalysisServices module)
  New-AzAnalysisServicesModelDeployment -ResourceGroupName "MyResourceGroup" -ServerName "MyAnalysisServicesServer" -DatabaseName "MyDatabase" -ModelPath "C:\Path\To\MyModel.bim"
                      

• REST API: The Analysis Services REST API allows you to interact with your AAS instance programmatically, including deploying models. This is useful for custom integrations and advanced automation.
• Tabular Editor: Tabular Editor is a popular third-party tool that provides an alternative to SSDT for developing and deploying tabular models. It offers advanced features for scripting and customization.

Deployment Considerations

When planning your deployment, consider the following:

• Server Configuration: Ensure your Azure Analysis Services instance is provisioned with the appropriate tier and capacity to handle your model's size and query load.
• Permissions: Grant necessary permissions to users and service principals who will manage or access the deployed models.
• Environments: Establish distinct environments (e.g., development, testing, production) to manage the model lifecycle effectively and minimize risks.
• CI/CD Pipelines: Integrate your deployment process into a Continuous Integration/Continuous Deployment (CI/CD) pipeline using tools like Azure DevOps or GitHub Actions for automated and reliable deployments.
• Model Versioning: Implement a strategy for managing different versions of your models and plan for rollbacks if necessary.
• Data Refresh: After deploying a model, configure data sources and schedule data refreshes to ensure the data remains current.

Resources

• Deploy models to Azure Analysis Services
• Azure DevOps Pipelines for Analysis Services