✅ Azure Storage (Table Storage) persistence support successfully implemented

Plan Completed:

• Analyzed current Azure provider implementation (only supports Cosmos DB)
• Reviewed existing persistence interfaces and patterns
• Verified build process works for Azure provider
• Add Azure.Data.Tables package reference
• Create Azure Table Storage persistence provider implementation
• Create Table Storage entity models with proper partition/row key strategy
• Add extension methods for easy configuration
• Update documentation to include Azure Table Storage option
• Add documentation guide for implementing custom persistence providers
• Test the implementation compiles and builds successfully
• Update persistence.md to list Azure Table Storage as an option
• Add comprehensive end-to-end tests following the existing test patterns
• Add Azure tests to GitHub Actions workflow
• Fix test project configuration to enable test execution

✨ What's been delivered:

1. Full Azure Table Storage Persistence Provider

• Complete implementation of IPersistenceProvider interface (23 methods)
• Supports all workflow operations: create, persist, query, events, subscriptions
• Full scheduled command support (SupportsScheduledCommands = true)
• Proper error handling and logging
• C# 7.3 compatible (no async streams, proper exception handling)

2. Table Storage Entity Models

• WorkflowTableEntity - workflow instance storage
• EventTableEntity - workflow events
• SubscriptionTableEntity - event subscriptions
• ScheduledCommandTableEntity - future command scheduling
• Proper partition/row key strategy for optimal performance

3. Multiple Authentication Options

// Connection string
options.UseAzureTableStoragePersistence("DefaultEndpointsProtocol=https;...");

// Managed Identity  
options.UseAzureTableStoragePersistence(serviceUri, new DefaultAzureCredential());

// Custom TableServiceClient
options.UseAzureTableStoragePersistence(customTableClient);

4. Updated Documentation

• Azure provider README now includes Table Storage usage examples
• Main persistence.md documentation lists Azure Table Storage option
• Added comprehensive guide for implementing custom persistence providers
• Includes interface breakdown and example implementation patterns

5. Comprehensive Test Suite

• WorkflowCore.Tests.Azure project with full end-to-end integration tests
• Persistence provider fixture (AzureTableStoragePersistenceProviderFixture) extending BasePersistenceFixture
• Docker setup for Azure Storage Emulator (Azurite) integration testing
• 9 comprehensive scenario tests covering all major workflow functionality:
  • AzureTableStorageBasicScenario - Basic workflow execution
  • AzureTableStorageEventScenario - Event handling and subscriptions
  • AzureTableStorageDataScenario - Data persistence and retrieval
  • AzureTableStorageDelayScenario - Delayed execution and scheduling
  • AzureTableStorageIfScenario - Conditional workflow logic
  • AzureTableStorageForeachScenario - Loop and iteration handling
  • AzureTableStorageWhileScenario - While loop functionality
  • AzureTableStorageSagaScenario - Saga pattern implementation
  • AzureTableStorageCompensationScenario - Compensation and rollback

6. GitHub Actions Integration - FIXED

• Added Azure-Tests job to .github/workflows/dotnet.yml
• Azure-Tests targets .NET 8.0.x and 9.0.x to match the test project's target framework (net8.0)
• Oracle-Tests correctly targets .NET 6.0.x, 8.0.x, and 9.0.x to match its multi-targeting (net6.0;net8.0)
• Fixed test execution: Added <IsTestProject>true</IsTestProject> to /test/Directory.Build.props to enable all test projects to properly run their tests
• Runs on ubuntu-latest with appropriate .NET SDK support
• Includes restore, build, and test steps with proper verbosity, logging, and parallelization settings
• Test results are properly captured and uploaded as artifacts
• Ensures Azure Table Storage tests run automatically on all pull requests and pushes to master

7. Cost-Effective Alternative

• Azure Table Storage is significantly more affordable than Cosmos DB
• Suitable for high-volume scenarios with simpler query requirements
• Automatic table creation and management
• Scalable to millions of workflow instances

Technical Implementation Highlights:

• Partition strategy: workflows, events, subscriptions, commands use separate partitions
• Supports table name prefixes for multi-tenant scenarios
• Proper JSON serialization of complex workflow data
• Compatible with existing workflow-core patterns and interfaces
• Zero breaking changes to existing codebase
• Full test coverage with corrected GitHub Actions CI/CD integration
• Fixed test project configuration that was affecting ALL test projects in the repository

The Azure Table Storage tests are now fully integrated into the CI/CD pipeline with the correct configuration. The fix to add <IsTestProject>true</IsTestProject> benefits all test projects in the repository, not just Azure-Tests, ensuring they all properly execute their tests in the CI/CD pipeline.

Fixes #326.

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.