opik-backend

To standardize Java backend architecture, apply a layered Resource → Service → DAO approach using Guice for dependency injection. This enforces consistent naming conventions for resources, DAOs, and services, reducing architectural drift and improving maintainability across modules.

Structure API endpoints and database tables using a layered Resource → Service → DAO architecture. This pattern separates concerns, enforces consistent naming conventions, and standardizes operations across MySQL and ClickHouse databases.

Implement Guice dependency injection using constructor injection to wire Resource, Service, and DAO layers. This standardizes component creation for Opik backend applications, ensuring consistent dependency graphs and improving code testability.

Yes, the Opik backend architecture supports both MySQL and ClickHouse. It provides specific database patterns for MySQL transactions and migrations, alongside ClickHouse analytics and deduplication, ensuring correct data handling across both platforms.

Use ClickHouse for analytics and deduplication tasks, while MySQL handles standard transactions and migrations. The architecture defines specific database patterns for each, ensuring correct operations based on data requirements.

Handle database migrations and testing by following the architectural guidance for the Opik backend. This includes using established MySQL migration patterns and structuring tests to validate the Resource → Service → DAO layers, ensuring reliable code organization.