5. Deploying an Aqueduct Application

The last chapter is a quick one - we'll get our application and its database running locally. When writing tests, the harness creates temporary tables that are destroyed when the tests end. Those tables are created in a database named dart_test that is exclusively used for this purpose. All of your projects will use this same database for running tests.

To run the application outside of the tests, you'll need another database. Run the psql command-line tool and enter the following SQL:

CREATE DATABASE quiz;
CREATE USER quiz_user WITH createdb;
ALTER USER quiz_user WITH password 'quizzy';
GRANT all ON database quiz TO quiz_user;

This creates a database quiz that quiz_user has access to. Now, add quiz's data model to this database by running the following commands in the project directory:

aqueduct db generate
aqueduct db upgrade --connect postgres://quiz_user:quizzy@localhost:5432/quiz

The first command generates a migration file in migrations/ that adds tables _Question and _Answer, and the second command executes that migration file on the newly created database.

After adding the data model to the quiz database, run the following commands in psql to insert a question and answer:

\c quiz
INSERT INTO _question (description) VALUES ('What is 1+1?');
INSERT INTO _answer (description, question_index) VALUES ('2', 1);

The application is currently hard-coded to connect to the test database. We'll write a bit of code to read connection info from a YAML configuration file instead. At the bottom of quiz_sink.dart, create a Configuration subclass:

class QuizConfig extends Configuration {
  QuizConfig(String filename) : super.fromFile(filename);

  DatabaseConnectionConfiguration database;
}

Update QuizSink's constructor to create its persistent store from configuration values:

QuizSink(ApplicationOptions appConfig) : super(appConfig) {
  logger.onRecord.listen((rec) =>
    print("$rec ${rec.error ?? ""} ${rec.stackTrace ?? ""}"));
  var dataModel = new ManagedDataModel.fromCurrentMirrorSystem();

  var configValues = new QuizConfig(appConfig.configurationFilePath);

  var persistentStore = new PostgreSQLPersistentStore.fromConnectionInfo(
      configValues.database.username,
      configValues.database.password,
      configValues.database.host,
      configValues.database.port,
      configValues.database.databaseName);
  context = new ManagedContext(dataModel, persistentStore);
}

Finally, create the file config.yaml in the root of the project directory and add the following key-values pairs:

database:
 username: quiz_user
 password: quizzy
 host: localhost
 port: 5432
 databaseName: quiz

Run aqueduct serve and open a browser to http://localhost:8888/questions - you'll see the question in your database. For other ways of running an Aqueduct application (and tips for running them remotely), see this guide.

Test and Deployment Configuration

The configurationFilePath defaults to config.yaml when using aqueduct serve. In the test harness, the configurationFilePath is set to config.src.yaml. To continue running the tests, add the database connection configuration for dart_test database to the file config.src.yaml. For more details on configuration, see this guide.

Onward

We've only touched on a small part of Aqueduct, but we've hit the fundamentals pretty well. The rest of the guides on this site will take you deeper on these topics, and topics we haven't covered like OAuth 2.0.

It's very important that you get comfortable using the API reference in addition to these guides. If you are looking to solve a problem, start by looking at the API reference for all of the objects you have access to (including the type you are writing the method for). The properties and methods you have access to will lead you to more properties and methods that'll eventually do what you want done.

Users of the documentation viewer Dash can add Aqueduct through the Preferences pane, under Downloads.

There are IntelliJ IDEA file and code templates available for Aqueduct. See this guide for installation instructions and usage. It takes 10 seconds and it'll save you a ton of time overall.

And lastly, remember to create a new project:

aqueduct create my_next_big_idea