Loading Test Data
Learn how to load data through the reactive repository, its drawbacks, and a solution to use a nonreactive CRUD repository.
We'll cover the following...
Let’s use the repository we just created to load some test data.
Using ItemRepository
If we wanted to grab a copy of our delightful ItemRepository, we’d probably write something like this:
itemRepository.save(new Item("Alf alarm clock", 19.99))
See the problem?
The ReactiveCrudRepository.save() method returns a Mono<T>. In this case, it’s Mono<Item>. Remember, a Mono Reactor type doesn’t actually do anything until we subscribe.
In short, the code above does absolutely nothing. To actually get things moving, we must do this:
itemRepository.save(new Item("Alf alarm clock", 19.99)).subscribe()
Keep in mind that there’s still a problem with this, at least during startup. When Netty is launched, the risk of a subscriber causing the startup thread to deadlock its event loop is very real.
Using a nonreactive CRUD repository
When attempting to do something during the startup of our application, the recommendation is to use a blocking version of Spring Data MongoDB!
Because this is about application startup, and focused on test configuration, it’s okay to allow a little bit of blocking code. That’s as long as it doesn’t escape into production!
The only question is, what are we blocking? The repository we defined earlier extends ReactiveCrudRepository. We can create a counterpart repository.
interface BlockingItemRepository extends CrudRepository<Item, String> {}
This repository definition is ...