Tell me more about your implementation of CQRS in your case.

In our project, CQRS (Command Query Responsibility Segregation) is implemented through the separation of data models and services for commands (changes) and queries (reads). A separate layer is used for commands, which handles business logic and changes the system state, for example, through services working with aggregates and repositories, ensuring consistency and validation.

For queries, a separate read-only layer is implemented, optimized for fast and scalable read operations. We use denormalized views or specialized databases as storage for reading, which allows quick data retrieval without complex join operations.

An asynchronous synchronization between the command and query layers is implemented via events (Event Sourcing or simple event publishing), which ensures eventual consistency. This allows scaling read and write operations independently.

An example of a simplified structure:

// Command
public class CreateOrderCommand {
    private String orderId;
    private List<Item> items;
    // getters and setters
}

// Command handler
public class OrderCommandHandler {
    public void handle(CreateOrderCommand command) {
        // validation, business logic
        orderRepository.save(new Order(command.getOrderId(), command.getItems()));
        eventBus.publish(new OrderCreatedEvent(command.getOrderId()));
    }
}

// Query
public class OrderQueryService {
    public OrderDto getOrderById(String orderId) {
        return orderReadRepository.findById(orderId);
    }
}

Thus, CQRS helps us improve performance and simplify the maintenance of complex business logic.