# Concurrency in Dart

> Use isolates to enable parallel code execution on multiple processor cores.



<?code-excerpt path-base="concurrency"?>

This page contains a conceptual overview of how concurrent programming works in
Dart. It explains the event-loop, async language features, and isolates from
a high-level. For more practical code examples of using concurrency in Dart,
read the [Asynchronous programming](/language/async) page and 
[Isolates](/language/isolates) page.

Concurrent programming in Dart refers to both asynchronous APIs, like `Future`
and `Stream`, and *isolates*, which allow you to move processes to separate
cores.

All Dart code runs in isolates, starting in the default main isolate,
and optionally expanding to whatever subsequent isolates you
explicitly create. When you spawn a new isolate,
it has its own isolated memory, and its own event loop.
The event loop is what makes asynchronous and
concurrent programming possible in Dart.

## Event Loop

Dart’s runtime model is based on an event loop.
The event loop is responsible for executing your program's code,
collecting and processing events, and more.

As your application runs, all events are added to a queue,
called the *event queue*.
Events can be anything from requests to repaint the UI,
to user taps and keystrokes, to I/O from the disk.
Because your app can’t predict what order events will happen,
the event loop processes events in the order they're queued, one at a time.

![A figure showing events being fed, one by one, into the event loop](/assets/img/language/concurrency/event-loop.png){:.diagram-wrap}

The way the event loop functions resembles this code:

```dart
while (eventQueue.waitForEvent()) {
  eventQueue.processNextEvent();
}
```

This example event loop is synchronous and runs on a single thread.
However, most Dart applications need to do more than one thing at a time. 
For example, a client application might need to execute an HTTP request, 
while also listening for a user to tap a button. 
To handle this, Dart offers many async APIs, 
like [Futures, Streams, and async-await](/language/async).
These APIs are built around this event loop.

For example, consider making a network request:

```dart
http.get('https://example.com').then((response) {
  if (response.statusCode == 200) {
    print('Success!');
  }  
}
```

When this code reaches the event loop, it immediately calls the
first clause, `http.get`, and returns a `Future`.
It also tells the event loop to hold onto the callback in the `then()` clause
until the HTTP request resolves. When that happens, it should
execute that callback, passing the result of the request as an argument.

![Figure showing async events being added to an event loop and holding onto a callback to execute later .](/assets/img/language/concurrency/async-event-loop.png){:.diagram-wrap}

This same model is generally how the event loop handles all other
asynchronous events in Dart, such as [`Stream`][] objects.

[`Stream`]: https://api.dart.dev/dart-async/Stream-class.html

## Asynchronous programming

This section summarizes the different types and syntaxes of asynchronous programming in Dart.
If you're already familiar with `Future`, `Stream`, and async-await,
then you can skip ahead to the [isolates section][].

[isolates section]: #isolates

### Futures

A `Future` represents the result of an asynchronous operation that will 
eventually complete with a value or an error.

In this sample code, the return type of `Future<String>` represents a 
promise to eventually provide a `String` value (or error).

<?code-excerpt "lib/future_syntax.dart (read-async)"?>
```dart
Future<String> _readFileAsync(String filename) {
  final file = File(filename);

  // .readAsString() returns a Future.
  // .then() registers a callback to be executed when `readAsString` resolves.
  return file.readAsString().then((contents) {
    return contents.trim();
  });
}
```

### The async-await syntax

The `async` and `await` keywords provide a declarative way to define
asynchronous functions and use their results.

Here's an example of some synchronous code
that blocks while waiting for file I/O:

<?code-excerpt "lib/sync_number_of_keys.dart (blocking)"?>
```dart
const String filename = 'with_keys.json';

void main() {
  // Read some data.
  final fileData = _readFileSync();
  final jsonData = jsonDecode(fileData);

  // Use that data.
  print('Number of JSON keys: ${jsonData.length}');
}

String _readFileSync() {
  final file = File(filename);
  final contents = file.readAsStringSync();
  return contents.trim();
}
```

Here's similar code, but with changes (highlighted) to make it asynchronous:

<?code-excerpt "lib/async_number_of_keys.dart (non-blocking)" replace="/async|await|readAsString\(\)/[!$&!]/g; /Future<\w+\W/[!$&!]/g;"?>
```dart
const String filename = 'with_keys.json';

void main() [!async!] {
  // Read some data.
  final fileData = [!await!] _readFileAsync();
  final jsonData = jsonDecode(fileData);

  // Use that data.
  print('Number of JSON keys: ${jsonData.length}');
}

[!Future<String>!] _readFileAsync() [!async!] {
  final file = File(filename);
  final contents = [!await!] file.[!readAsString()!];
  return contents.trim();
}
```

The `main()` function uses the `await` keyword in front of `_readFileAsync()`
to let other Dart code (such as event handlers) use the CPU while native code
(file I/O) executes. Using `await` also has the effect of converting
the `Future<String>` returned by `_readFileAsync()` into a `String`. As a
result, the `contents` variable has the implicit type `String`.

:::note
The `await` keyword works only in functions that have `async` before the
function body.
:::

As the following figure shows, the Dart code pauses while `readAsString()`
executes non-Dart code, in either the Dart runtime or the operating system. 
Once `readAsString()` returns a value, Dart code execution resumes.

![Flowchart-like figure showing app code executing from start to exit, waiting for native I/O in between](/assets/img/language/concurrency/basics-await.png){:.diagram-wrap}

### Streams

Dart also supports asynchronous code in the form of streams. Streams
provide values in the future and repeatedly over time. A promise to provide a
series of `int` values over time has the type `Stream<int>`.

In the following example, the stream created with `Stream.periodic`
repeatedly emits a new `int` value every second.

<?code-excerpt "lib/stream_syntax.dart"?>
```dart
Stream<int> stream = Stream.periodic(const Duration(seconds: 1), (i) => i * i);
```

#### await-for and yield

Await-for is a type of for loop that executes each subsequent iteration of the
loop as new values are provided. In other words, it’s used to “loop over”
streams. In this example, a new value will be emitted from the function
`sumStream` as new values are emitted from the stream that’s provided as an
argument. The `yield` keyword is used rather than `return` in functions that 
return streams of values.

<?code-excerpt "lib/await_for_syntax.dart"?>
```dart
Stream<int> sumStream(Stream<int> stream) async* {
  var sum = 0;
  await for (final value in stream) {
    yield sum += value;
  }
}
```

If you'd like to learn more about using `async`, `await`, `Stream`s and
`Future`s, check out the [asynchronous programming tutorial][].

[asynchronous programming tutorial]: /libraries/async/async-await

## Isolates

Dart supports concurrency via isolates, in addition to [asynchronous APIs](#asynchronous-programming).
Most modern devices have multi-core
CPUs. To take advantage of multiple cores, developers sometimes use
shared-memory threads running concurrently. However, shared-state concurrency is
[error prone](https://en.wikipedia.org/wiki/Race_condition#In_software) and can
lead to complicated code.

Instead of threads, all Dart code runs inside isolates.
Using isolates, your Dart code can perform multiple independent tasks at once,
using additional processor cores if they're available.
Isolates are like threads or processes, but each isolate has its own memory
and a single thread running an event loop.

Each isolate has its own global fields,
ensuring that none of the state in an isolate is accessible
from any other isolate. Isolates can only communicate to each other via message
passing. No shared state between isolates means concurrency complexities like
[mutexes or locks](https://en.wikipedia.org/wiki/Lock_(computer_science))
and [data races](https://en.wikipedia.org/wiki/Race_condition#Data_race)
won't occur in Dart. That said, isolates don't prevent race conditions all
together. For more information on this concurrency model, read about
the [Actor model](https://en.wikipedia.org/wiki/Actor_model).

:::note Platform note
Only the [Dart Native platform][] implements isolates.
To learn more about the Dart Web platform,
see the [Concurrency on the web](#concurrency-on-the-web) section.
:::

[Dart Native platform]: /overview#platform

### The main isolate

In most cases, you don't need to think about isolates at all. Dart programs run
in the main isolate by default. It’s the thread where a program starts to run
and execute, as shown in the following figure:

![A figure showing a main isolate, which runs `main()`, responds to events, and then exits](/assets/img/language/concurrency/basics-main-isolate.png){:.diagram-wrap}

Even single-isolate programs can execute smoothly. Before continuing to the next
line of code, these apps use [async-await][] to wait for asynchronous operations to
complete. A well-behaved app starts quickly, getting to the event loop as soon
as possible. The app then responds to each queued event promptly, using
asynchronous operations as necessary.

[async-await]: /libraries/async/async-await

### The isolate life cycle

As the following figure shows,
every isolate starts by running some Dart code,
such as the `main()` function.
This Dart code might register some event listeners—to
respond to user input or file I/O, for example.
When the isolate's initial function returns,
the isolate stays around if it needs to handle events.
After handling the events, the isolate exits.

![A more general figure showing that any isolate runs some code, optionally responds to events, and then exits](/assets/img/language/concurrency/basics-isolate.png){:.diagram-wrap}

### Event handling

In a client app, the main isolate's event queue might contain repaint requests
and notifications of tap and other UI events. For example, the following figure
shows a repaint event, followed by a tap event, followed by two repaint events.
The event loop takes events from the queue in first in, first out order.

![A figure showing events being fed, one by one, into the event loop](/assets/img/language/concurrency/event-loop.png){:.diagram-wrap}

Event handling happens on the main isolate after `main()` exits. In the
following figure, after `main()` exits, the main isolate handles the first
repaint event. After that, the main isolate handles the tap event, followed by a
repaint event.

If a synchronous operation takes too much processing time, the app can become
unresponsive. In the following figure, the tap-handling code takes too long, so
subsequent events are handled too late. The app might appear to freeze, and any
animation it performs might be jerky.

![A figure showing a tap handler with a too-long execution time](/assets/img/language/concurrency/event-jank.png){:.diagram-wrap}

In client apps, the result of a too-lengthy synchronous operation is often
[janky (non-smooth) UI animation][jank]. Worse, the UI might become completely
unresponsive.

[jank]: https://docs.flutter.dev/perf/rendering-performance

### Background workers

If your app's UI becomes unresponsive due to a time-consuming
computation—[parsing a large JSON file][json], for example—consider offloading
that computation to a worker isolate, often called a _background worker._
A common case, shown in the following figure, is spawning a simple worker
isolate that performs a computation and then exits. The worker isolate returns
its result in a message when it exits.

[json]: https://docs.flutter.dev/cookbook/networking/background-parsing

![A figure showing a main isolate and a simple worker isolate](/assets/img/language/concurrency/isolate-bg-worker.png){:.diagram-wrap}

A worker isolate can perform I/O
(reading and writing files, for example), set timers, and more. It has its own
memory and doesn't share any state with the main isolate. The worker isolate can
block without affecting other isolates.

### Using isolates

There are two ways to work with isolates in Dart, depending on the use-case:

* Use [`Isolate.run()`][] to perform a single computation on a separate thread.
* Use [`Isolate.spawn()`][] to create an isolate that will handle multiple
  messages over time, or a background worker. For more information on working
  with long-lived isolates, read the [Isolates](/language/isolates) page.

In most cases, `Isolate.run` is the recommended
API to run processes in the background.

#### `Isolate.run()`

The static `Isolate.run()` method requires one argument: a callback that will be
run on the newly spawned isolate.

<?code-excerpt "lib/isolate_run_syntax.dart (slow)"?>
```dart
int slowFib(int n) => n <= 1 ? 1 : slowFib(n - 1) + slowFib(n - 2);

// Compute without blocking current isolate.
void fib40() async {
  var result = await Isolate.run(() => slowFib(40));
  print('Fib(40) = $result');
}
```

### Performance and isolate groups

When an isolate calls [`Isolate.spawn()`][], the two isolates have the same
executable code and are in the same _isolate group_. Isolate groups enable
performance optimizations such as sharing code; a new isolate immediately runs
the code owned by the isolate group. Also, `Isolate.exit()` works only when the
isolates are in the same isolate group.

In some special cases, you might need to use [`Isolate.spawnUri()`][], which
sets up the new isolate with a copy of the code that's at the specified URI.
However, `spawnUri()` is much slower than `spawn()`, and the new isolate isn't
in its spawner's isolate group. Another performance consequence is that message
passing is slower when isolates are in different groups.

[`Isolate.spawnUri()`]: https://api.dart.dev/dart-isolate/Isolate/spawnUri.html

### Limitations of isolates

#### Isolates aren't threads

If you’re coming to Dart from a language with multithreading, it’d be reasonable
to expect isolates to behave like threads, but that isn’t the case. Each isolate
has its own state, ensuring that none of the state in an isolate is
accessible from any other isolate. Therefore, isolates are limited by their
access to their own memory.

For example, if you have an application with a
global mutable variable, that variable will be a separate
variable in your spawned isolate. If you mutate that variable in the spawned
isolate, it will remain untouched in the main isolate. This is how isolates are
meant to function, and it's important to keep in mind when you’re considering
using isolates.

#### Message types

Messages sent via [`SendPort`][]
can be almost any type of Dart object, but there are a few exceptions:

- Objects with native resources, such as [`Socket`][].
- [`ReceivePort`][]
- [`DynamicLibrary`][]
- [`Finalizable`][]
- [`Finalizer`][]
- [`NativeFinalizer`][]
- [`Pointer`][]
- [`UserTag`][]
- Instances of classes that are marked with `@pragma('vm:isolate-unsendable')`

Apart from those exceptions, any object can be sent.
Check out the [`SendPort.send`][] documentation for more information.

Note that `Isolate.spawn()` and `Isolate.exit()` abstract over `SendPort` 
objects, so they're subject to the same limitations.

[`SendPort.send`]: https://api.dart.dev/dart-isolate/SendPort/send.html
[`Socket`]: https://api.dart.dev/dart-io/Socket-class.html
[`DynamicLibrary`]: https://api.dart.dev/dart-ffi/DynamicLibrary-class.html
[`Finalizable`]: https://api.dart.dev/dart-ffi/Finalizable-class.html
[`Finalizer`]: https://api.dart.dev/dart-core/Finalizer-class.html
[`NativeFinalizer`]: https://api.dart.dev/dart-ffi/NativeFinalizer-class.html
[`Pointer`]: https://api.dart.dev/dart-ffi/Pointer-class.html
[`UserTag`]: https://api.dart.dev/dart-developer/UserTag-class.html

#### Synchronous blocking communication between isolates

There is a limit to the number of isolates that can run in parallel.
This limit doesn't affect the standard *asynchronous* communication between isolates
via messages in Dart. You can have hundreds of isolates running concurrently
and making progress. The isolates are scheduled on the CPU in round-robin fashion,
and yield to each other often.

Isolates can only communicate *synchronously* outside of pure Dart,
using C code via [FFI] to do so. 
Attempting synchronous communication between isolates
by synchronous blocking in FFI calls
might result in deadlock if the number of isolates is over the limit,
unless special care is taken.
The limit is not hardcoded to a particular number,
it's calculated based on the Dart VM heap size available to the Dart application.

To avoid this situation, the C code performing synchronous blocking
needs to leave the current isolate before performing the blocking operation
and re-enter it before returning to Dart from the FFI call.
Read about [`Dart_EnterIsolate`] and [`Dart_ExitIsolate`] to learn more.

[FFI]: /interop/c-interop
[`Dart_EnterIsolate`]: https://github.com/dart-lang/sdk/blob/c9a8bbd8d6024e419b5e5f26b5131285eb19cc93/runtime/include/dart_api.h#L1254
[`Dart_ExitIsolate`]: https://github.com/dart-lang/sdk/blob/c9a8bbd8d6024e419b5e5f26b5131285eb19cc93/runtime/include/dart_api.h#L1455

<a id="web"></a>
## Concurrency on the web

All Dart apps can use `async-await`, `Future`, and `Stream`
for non-blocking, interleaved computations. The [Dart web platform][], however,
does not support isolates. Dart web apps can use [web workers][] to run scripts
in background threads similar to isolates. Web workers' functionality and
capabilities differ somewhat from isolates, though.

For instance, when web workers send data between threads, they copy the data
back and forth. Data copying can be very slow, though, especially for large
messages. Isolates do the same, but also provide APIs that can more efficiently
_transfer_
the memory that holds the message instead.

Creating web workers and isolates also differs. You can only create web workers
by declaring a separate program entrypoint and compiling it separately. Starting
a web worker is similar to using `Isolate.spawnUri` to start an isolate. You can
also start an isolate with `Isolate.spawn`, which requires fewer resources
because it
[reuses some of the same code and data](#performance-and-isolate-groups)
as the spawning isolate. Web workers don't have an equivalent API.

[Dart web platform]: /overview#platform
[web workers]: https://developer.mozilla.org/docs/Web/API/Web_Workers_API/Using_web_workers


## Additional resources

- If you’re using many isolates, consider
  the [`IsolateNameServer`][]
  in Flutter, or
  [`package:isolate_name_server`][] that provides
  similar functionality for non-Flutter Dart applications.
- Read more about [Actor model][], which Dart's isolates are based on.
- Additional documentation on `Isolate` APIs:
    - [`Isolate.exit()`][]
    - [`Isolate.spawn()`][]
    - [`ReceivePort`][]
    - [`SendPort`][]

[`IsolateNameServer`]: https://api.flutter.dev/flutter/dart-ui/IsolateNameServer-class.html
[`package:isolate_name_server`]: https://pub.dev/packages/isolate_name_server
[Actor model]: https://en.wikipedia.org/wiki/Actor_model
[`Isolate.run()`]: https://api.dart.dev/dart-isolate/Isolate/run.html
[`Isolate.exit()`]: https://api.dart.dev/dart-isolate/Isolate/exit.html
[`Isolate.spawn()`]: https://api.dart.dev/dart-isolate/Isolate/spawn.html
[`ReceivePort`]: https://api.dart.dev/dart-isolate/ReceivePort-class.html
[`SendPort`]: https://api.dart.dev/dart-isolate/SendPort-class.html

