MongoDB is a document-oriented database. It uses the concept of the document to store data, which is more flexible than the row concept in the relational database management system (RDBMS).
A document allows you to represent complex hierarchical relationships with a single record.
MongoDB doesn't require predefined schemas that allow you to add to or remove fields from documents more quickly.
When the database grows, you'll have a challenge of how to scale it. There are two common ways:
- Scaling up – upgrade the current server to a bigger one with more resources (CPU, RAM, etc). However, getting a bigger server means increasing more costs.
- Scaling out – purchase additional servers and add them to the cluster. This is cheaper and more scalable than scaling up. The downside is that it takes more effort to manage multiple servers than a big one.
MongoDB was designed to scale out, and it allows you to split data across many servers. It also automatically manages the load balancing across the cluster, redistributing data, and routing updates to the correct servers.
Like any database system, MongoDB allows you to insert, update, and delete, and select data. In addition, it supports other features including:
- Indexing
- Aggregation
- Specify collection and index types
- File Storage
Reactive programming is a programming paradigm that is functional, event-based, non-blocking, asynchronous, and centered around data stream processing. The term reactive comes from the fact that we react to changes such as mouse clicks or I/O events.
Reactive applications scale better and are more efficient when we are dealing with lots of streaming data. Reactive applications are non-blocking; they're not using resources waiting for processes to finish.
When building a reactive application, we need it to be reactive all the way down to the database. We need to use a database that supports reactive programming. MongoDB is a database that has reactive support.
Reactive applications implement an event-based model where data is pushed to the consumer. The consumer of data is called a subscriber, because it subscribes to the publisher, which publishes asynchronous streams of data.
Spring Reactor is a reactive library for building non-blocking applications on the JVM based on the Reactive Streams Specification. It offers two types of publishers: Mono and Flux. Flux is a publisher that produces 0 to N values. Operations that return multiple elements use this type. Mono is a publisher that produces 0 to 1 value. It is used for operations that return a single element. In the following application, we use reactive programming with a MongoDB database.
run this command to start all services in the correct order.
$ docker-compose up -d$ docker exec -it mongodb bash$ mongosh "mongodb://localhost:27017/spring" --username root --password password --authenticationDatabase admin$ db.getMongo()$ db.getCollectionNames()$ db.users.insert({"firstName" : "John",
"lastName" : "Doe",
"userName": "john",
"password": "password"})$ db.users.find()docker-compose down