Get started

This page will help you understand how a Serverpod project is structured, how to make calls to endpoints, and how to communicate with the database.

Serverpod or Serverpod Mini?

Serverpod Mini is a lightweight version of Serverpod that is perfect for small projects or when you want to try out Serverpod without setting up a Postgres database. If you start with Mini, you can upgrade to the full version of Serverpod anytime.

Get started with Mini

Serverpod vs Serverpod Mini comparison

Feature	Serverpod	Serverpod Mini
Remote method calls	✅	✅
Generated data models	✅	✅
Streaming data	✅	✅
Custom auth	✅	✅
Pre-built auth	✅
Postgres database ORM	✅
Task scheduling	✅
Basic logging	✅	✅
Serverpod Insights	✅
Caching	✅	✅
File uploads	✅
Health checks	✅
Relic web server	✅
Easy deployment	✅	✅

Creating a new Serverpod project

The full version of Serverpod needs access to a Postgres database. The easiest way to set that up is to use our pre-configured Docker container. Install Flutter, Serverpod and Docker Desktop before you begin.

Create a new project by running serverpod create.

$ serverpod create mypod

info

Serverpod executes the flutter create command inside the flutter package during project creation. On Windows, flutter commands require that developer mode is enabled in the system settings.

This command will create a new directory called mypod, with three dart packages inside; mypod_server, mypod_client, and mypod_flutter.

• mypod_server: This package contains your server-side code. Modify it to add new endpoints or other features your server needs.
• mypod_client: This is the code needed to communicate with the server. Typically, all code in this package is generated automatically, and you should not edit the files in this package.
• mypod_flutter: This is the Flutter app, pre-configured to connect to your local server.

Starting the server

Make sure that Docker Desktop is running, then start your Docker containers with docker compose up --build --detach. It will start Postgres and Redis. Then, run dart bin/main.dart --apply-migrations to start your server.

$ cd mypod/mypod_server
$ docker compose up --build --detach
$ dart bin/main.dart --apply-migrations

If everything is working, you should see something like this on your terminal:

SERVERPOD version: 2.x.x, mode: development, time: 2022-09-12 17:22:02.825468Z
Insights listening on port 8081
Server default listening on port 8080
Webserver listening on port 8082

info

If you need to stop the Docker containers at some point, just run docker compose stop or use the Docker Desktop application. You can also use Docker Desktop to start, stop, and manage your containers.

important

In your development environment it can be helpful to always start Serverpod with the --apply-migrations flag, as this will ensure that the database is always up-to-date with your latest migration. However, in production you should typically start the server without the flag, unless you want to actually apply a new migration.

Running the demo app

Start the default demo app by changing the directory into the Flutter package that was created and running flutter run.

$ cd mypod/mypod_flutter
$ flutter run -d chrome

The flag -d chrome runs the app in Chrome, for other run options please see the Flutter documentation.

info

iOS Simulator: Because an iOS simulator has its own localhost, it won't find the server running on your machine. Therefore, you will need to pass the IP address of your machine when creating the client in mypod/mypod_flutter/lib/main.dart. Depending on your local network, it might look something like this:

var client = Client('http://192.168.1.117:8080/')
  ..connectivityMonitor = FlutterConnectivityMonitor();

info

MacOS: If you run the app on MacOS, you will need to add permissions for outgoing connections in your Xcode project. To do this, open the Runner.xcworkspace in Xcode. Then check the Outgoing Connections (Client) under Runner > Signing & Capabilities > App Sandbox. Make sure to add the capability for all run configurations.

Server overview

At first glance, the complexity of the server may seem daunting, but there are only a few directories and files you need to pay attention to. The rest of the files will be there when you need them in the future, e.g., when you want to deploy your server or if you want to set up continuous integration.

These are the most important directories:

• config: These are the configuration files for your Serverpod. These include a password.yaml file with your passwords and configurations for running your server in development, staging, and production. By default, everything is correctly configured to run your server locally.
• lib/src/endpoints: This is the default location for your server's endpoints. When you add methods to an endpoint, Serverpod will generate the corresponding methods in your client.
• lib/src/models: Default location for your model definition files. The files define the classes you can pass through your API and how they relate to your database. Serverpod generates serializable objects from the model definitions.

Both the endpoints and models directories contain sample files that give a quick idea of how they work. So this a great place to start learning.

Generating code

Whenever you change your code in either the endpoints or models directory, you will need to regenerate the classes managed by Serverpod. Do this by running serverpod generate.

$ cd mypod/mypod_server
$ serverpod generate

Working with endpoints

Endpoints are the connection points to the server from the client. With Serverpod, you add methods to your endpoint, and your client code will be generated. For the code to be generated, you need to place your endpoint in the lib/src/endpoints directory of your server. Your endpoint should extend the Endpoint class. For methods to be generated, they need to return a typed Future, and its first parameter should be a Session object. The Session object holds information about the call being made and provides access to the database.

import 'package:serverpod/serverpod.dart';

class ExampleEndpoint extends Endpoint {
  Future<String> hello(Session session, String name) async {
    return 'Hello $name';
  }
}

The above code will create an endpoint called example (the Endpoint suffix will be removed) with the single hello method. To generate the client-side code run serverpod generate in the home directory of the server.

On the client side, you can now invoke the method by calling:

var result = await client.example.hello('World');

tip

To learn more about endpoints, see the Working with endpoints section.

Serializing data

Serverpod makes it easy to generate serializable classes that can be passed between server and client or used to communicate with the database.

The structure for your serialized classes is defined in .spy.yaml files anywhere in the lib directory. Run serverpod generate in the home directory of the server to build the Dart code for the classes and make them accessible to both the server and client.

Here is a simple example of a .spy.yaml file defining a serializable class:

class: Company
fields:
  name: String
  foundedDate: DateTime?

Supported types are bool, int, double, String, DateTime, ByteData, and other serializable classes. You can also use Lists and Maps of the supported types, just make sure to specify the types. Null safety is supported. The keys of Map must be non-nullable Strings. Once your classes are generated, you can use them as parameters or return types to endpoint methods.

tip

You can also create custom serialized classes with tools such as Freezed. Learn more in the Serialization section.

Working with the database

A core feature of Serverpod is to query the database easily. Serverpod provides an ORM that supports type and null safety.

Connecting to the database

When working with the database, it is common that you want to connect to it with a database viewer such as Postico2, PgAdmin, or DBeaver. To connect to the database, you need to specify the host and port along with the database name, user name, and password. In your project, you can find these inside the config directory.

The connection details can be found in the file config/development.yaml. The variable name refers to the database name (which is your project name only).

database:
  host: localhost
  port: 8090
  name: projectname
  user: postgres

...

The password can be found in the file config/passwords.yaml.

development:
  database: '<MY DATABASE PASSWORD>'

...

Migrations

With database migrations, Serverpod makes it easy to evolve your database schema. When you make changes to your project that should be reflected in your database, you need to create a migration. A migration is a set of SQL queries that are run to update the database. To create a migration, run serverpod create-migration in the home directory of the server.

$ cd mypod/mypod_server
$ serverpod create-migration

Migrations are then applied to the database as part of the server startup by adding the --apply-migrations flag.

$ cd mypod/mypod_server
$ dart bin/main.dart --apply-migrations

tip

To learn more about database migrations, see the Migrations section.

Object database mapping

Add a table key to your model file to add a mapping to the database. The value specified after the key sets the database table name. Here is the Company class from earlier with a database table mapping to a table called company:

class: Company
table: company
fields:
  name: String
  foundedDate: DateTime?

CRUD operations are available through the static db method on all classes with database bindings.

tip

To learn more about database CRUD operations, see the CRUD section.

Writing to database

Inserting a new row into the database is as simple as calling the static db.insertRow method.

var myCompany = Company(name: 'Serverpod corp.', foundedDate: DateTime.now());
myCompany = await Company.db.insertRow(session, myCompany);

The method returns the inserted object with its id field set from the database.

Reading from database

Retrieving a single row from the database can done by calling the static db.findById method and providing the id of the row.

var myCompany = await Company.db.findById(session, companyId);

You can also use an expression to do a more refined search through the db.findFirstRow(...). method. The where parameter is a typed expression builder. The builder's parameter, t, contains a description of the table and gives access to the table's columns.

var myCompany = await Company.db.findFirstRow(
  session,
  where: (t) => t.name.equals('My Company'),
);

The example above will return a single row from the database where the name column is equal to My Company.

If no matching row is found, null is returned.

tip

Working with a database is an extensive subject. Learn more in the Database section.

Where to go next

You should now have a basic understanding of how Serverpod works. The different topics are described in more detail in the Concepts section of the documentation. If you are unfamiliar with server-side development, a good starting place for learning is to do the Build your first app tutorial. There are also many good video tutorials linked in the Tutorials section.

If you get stuck, never be afraid to ask questions in our community on Github. The Serverpod team is very active there, and many questions are also answered by other developers in the community.