Skip to main content

Core application - .NET SDK

This page shows how to do the following:

Develop a Workflow

How to develop a basic Workflow using the Temporal .NET SDK

Workflows are the fundamental unit of a Temporal Application, and it all starts with the development of a Workflow Definition.

In the Temporal .NET SDK programming model, Workflows are defined as classes.

Specify the [Workflow] attribute from the Temporalio.Workflows namespace on the Workflow class to identify a Workflow.

Use the [WorkflowRun] attribute to mark the entry point method to be invoked. This must be set on one asynchronous method defined on the same class as [Workflow].

using Temporalio.Workflows;

[Workflow]
public class MyWorkflow
{
public async Task<string> RunAsync(string name)
{
var param = MyActivityParams("Hello", name);
return await Workflow.ExecuteActivityAsync(
(MyActivities a) => a.MyActivity(param),
new() { StartToCloseTimeout = TimeSpan.FromMinutes(5) });
}
}

Temporal Workflows may have any number of custom parameters. However, we strongly recommend that objects are used as parameters, so that the object's individual fields may be altered without breaking the signature of the Workflow. All Workflow Definition parameters must be serializable.

Workflow logic requirements

Workflow logic is constrained by deterministic execution requirements. Therefore, each language is limited to the use of certain idiomatic techniques. However, each Temporal SDK provides a set of APIs that can be used inside your Workflow to interact with external (to the Workflow) application code.

This means there are several things Workflows cannot do such as:

  • Perform IO (network, disk, stdio, etc)
  • Access/alter external mutable state
  • Do any threading
  • Do anything using the system clock (e.g. DateTime.Now)
    • This includes .NET timers (e.g. Task.Delay or Thread.Sleep)
  • Make any random calls
  • Make any not-guaranteed-deterministic calls (e.g. iterating over a dictionary)

.NET Task Determinism

Some calls in .NET do unsuspecting non-deterministic things and are easy to accidentally use. This is especially true with Tasks. Temporal requires that the deterministic TaskScheduler.Current is used, but many .NET async calls will use TaskScheduler.Default implicitly (and some analyzers even encourage this). Here are some known gotchas to avoid with .NET tasks inside of Workflows:

  • Do not use Task.Run - this uses the default scheduler and puts work on the thread pool.
    • Use Workflow.RunTaskAsync instead.
    • Can also use Task.Factory.StartNew with current scheduler or instantiate the Task and run Task.Start on it.
  • Do not use Task.ConfigureAwait(false) - this will not use the current context.
    • If you must use Task.ConfigureAwait, use Task.ConfigureAwait(true).
    • There is no significant performance benefit to Task.ConfigureAwait in workflows anyways due to how the scheduler works.
  • Do not use anything that defaults to the default task scheduler.
  • Do not use Task.Delay, Task.Wait, timeout-based CancellationTokenSource, or anything that uses .NET built-in timers.
    • Workflow.DelayAsync, Workflow.WaitConditionAsync, or non-timeout-based cancellation token source is suggested.
  • Do not use Task.WhenAny.
    • Use Workflow.WhenAnyAsync instead.
    • Technically this only applies to an enumerable set of tasks with results or more than 2 tasks with results. Other uses are safe. See this issue.
  • Do not use Task.WhenAll
    • Use Workflow.WhenAllAsync instead.
    • Technically Task.WhenAll is currently deterministic in .NET and safe, but it is better to use the wrapper to be sure.
  • Do not use CancellationTokenSource.CancelAsync.
    • Use CancellationTokenSource.Cancel instead.
  • Do not use System.Threading.Semaphore or System.Threading.SemaphoreSlim or System.Threading.Mutex.
    • Use Temporalio.Workflows.Semaphore or Temporalio.Workflows.Mutex instead.
    • Technically SemaphoreSlim does work if only the async form of WaitAsync is used without no timeouts and Release is used. But anything else can deadlock the workflow and its use is cumbersome since it must be disposed.
  • Be wary of additional libraries' implicit use of the default scheduler.
    • For example, while there are articles for Dataflow about using a specific scheduler, there are hidden implicit uses of TaskScheduler.Default. For example, see this bug.

In order to help catch wrong scheduler use, by default the Temporal .NET SDK adds an event source listener for info-level task events. While this technically receives events from all uses of tasks in the process, we make sure to ignore anything that is not running in a Workflow in a high performant way (basically one thread local check). For code that does run in a Workflow and accidentally starts a task in another scheduler, an InvalidWorkflowOperationException will be thrown which "pauses" the Workflow (fails the Workflow Rask which continually retries until the code is fixed). This is unfortunately a runtime-only check, but can help catch mistakes early. If this needs to be turned off for any reason, set DisableWorkflowTracingEventListener to true in Worker options.

In the near future for modern .NET versions we hope to use the new TimeProvider API which will allow us to control current time and timers.

Workflow .editorconfig

Since Workflow code follows some different logic rules than regular C# code, there are some common analyzer rules that developers may want to disable. To ensure these are only disabled for Workflows, current recommendation is to use the .workflow.cs extension for files containing Workflows.

Here are the rules to disable:

  • CA1024 - This encourages properties instead of methods that look like getters. However for reflection reasons we cannot use property getters for queries, so it is very normal to have

    [WorkflowQuery]
    public string GetSomeThing() => someThing;
  • CA1822 - This encourages static methods when methods don't access instance state. Workflows however use instance methods for run, Signals, Queries, or Updates even if they could be static.

  • CA2007 - This encourages users to use ConfigureAwait instead of directly waiting on a task. But in Workflows, there is no benefit to this and it just adds noise (and if used, needs to be ConfigureAwait(true) not ConfigureAwait(false)).

  • CA2008 - This encourages users to always apply an explicit task scheduler because the default of TaskScheduler.Current is bad. But for Workflows, the default of TaskScheduler.Current is good/required.

  • CA5394 - This discourages use of non-crypto random. But deterministic Workflows, via Workflow.Random intentionally provide a deterministic non-crypto random instance.

  • CS1998 - This discourages use of async on async methods that don't await. But Workflows handlers like Signals are often easier to write in one-line form this way, e.g. public async Task SignalSomethingAsync(string value) => this.value = value;.

  • VSTHRD105 - This is similar to CA2008 above in that use of implicit current scheduler is discouraged. That does not apply to Workflows where it is encouraged/required.

Here is the .editorconfig snippet for the above which may frequently change as more analyzers need to be adjusted:

##### Configuration specific for Temporal workflows #####
[*.workflow.cs]

# We use getters for queries, they cannot be properties
dotnet_diagnostic.CA1024.severity = none

# Don't force workflows to have static methods
dotnet_diagnostic.CA1822.severity = none

# Do not need ConfigureAwait for workflows
dotnet_diagnostic.CA2007.severity = none

# Do not need task scheduler for workflows
dotnet_diagnostic.CA2008.severity = none

# Workflow randomness is intentionally deterministic
dotnet_diagnostic.CA5394.severity = none

# Allow async methods to not have await in them
dotnet_diagnostic.CS1998.severity = none

# Don't avoid, but rather encourage things using TaskScheduler.Current in workflows
dotnet_diagnostic.VSTHRD105.severity = none

Customize Workflow Type

How to customize your Workflow Type name using the Temporal .NET SDK

Workflows have a Type that are referred to as the Workflow name.

The following examples demonstrate how to set a custom name for your Workflow Type.

You can customize the Workflow name with a custom name in the attribute. For example, [Workflow("my-workflow-name")]. If the name parameter is not specified, the Workflow name defaults to the unqualified class name.

using Temporalio.Workflows;

[Workflow("MyDifferentWorkflowName")]
public class MyWorkflow
{
public async Task<string> RunAsync(string name)
{
var param = MyActivityParams("Hello", name);
return await Workflow.ExecuteActivityAsync(
(MyActivities a) => a.MyActivity(param),
new() { StartToCloseTimeout = TimeSpan.FromMinutes(5) });
}
}

Develop an Activity

How to develop a basic Activity using the Temporal .NET SDK

One of the primary things that Workflows do is orchestrate the execution of Activities. An Activity is a normal method execution that's intended to execute a single, well-defined action (either short or long-running), such as querying a database, calling a third-party API, or transcoding a media file. An Activity can interact with world outside the Temporal Platform or use a Temporal Client to interact with a Temporal Service. For the Workflow to be able to execute the Activity, we must define the Activity Definition.

You can develop an Activity Definition by using the [Activity] attribute from the Temporalio.Activities namespace on the method. To register a method as an Activity with a custom name, use an attribute parameter, for example [Activity("your-activity")]. Otherwise, the activity name is the unqualified method name (sans an "Async" suffix if the method is async).

Activities can be asynchronous or synchronous.

using Temporalio.Activities;

public class MyActivities
{
// Activities can be async and/or static too. We just demonstrate instance methods since many
// use them that way.
[Activity]
public string MyActivity(MyActivityParams input) =>
$"{input.Greeting}, {input.Name}!";
}

There is no explicit limit to the total number of parameters that an Activity Definition may support. However, there is a limit to the total size of the data that ends up encoded into a gRPC message Payload.

A single argument is limited to a maximum size of 2 MB. And the total size of a gRPC message, which includes all the arguments, is limited to a maximum of 4 MB.

Also, keep in mind that all Payload data is recorded in the Workflow Execution Event History and large Event Histories can affect Worker performance. This is because the entire Event History could be transferred to a Worker Process with a Workflow Task.

Some SDKs require that you pass context objects, others do not. When it comes to your application data—that is, data that is serialized and encoded into a Payload—we recommend that you use a single object as an argument that wraps the application data passed to Activities. This is so that you can change what data is passed to the Activity without breaking a method signature.

Activity parameters are the method parameters of the method with the [Activity] attribute. These can be any data type Temporal can convert, including records. Technically this can be multiple parameters, but Temporal strongly encourages a single parameter containing all input fields.

Start Activity Execution

How to start an Activity Execution using the Temporal .NET SDK

Calls to spawn Activity Executions are written within a Workflow Definition. The call to spawn an Activity Execution generates the ScheduleActivityTask Command. This results in the set of three Activity Task related Events (ActivityTaskScheduled, ActivityTaskStarted, and ActivityTask[Closed])in your Workflow Execution Event History.

A single instance of the Activities implementation is shared across multiple simultaneous Activity invocations. Activity implementation code should be idempotent.

The values passed to Activities through invocation parameters or returned through a result value are recorded in the Execution history. The entire Execution history is transferred from the Temporal service to Workflow Workers when a Workflow state needs to recover. A large Execution history can thus adversely impact the performance of your Workflow.

Therefore, be mindful of the amount of data you transfer through Activity invocation parameters or Return Values. Otherwise, no additional limitations exist on Activity implementations.

To spawn an Activity Execution, use the ExecuteActivityAsync operation from within your Workflow Definition.

using Temporalio.Workflows;

[Workflow]
public class MyWorkflow
{
public async Task<string> RunAsync(string name)
{
var param = MyActivityParams("Hello", name);
return await Workflow.ExecuteActivityAsync(
(MyActivities a) => a.MyActivity(param),
new() { StartToCloseTimeout = TimeSpan.FromMinutes(5) });
}
}

Activity Execution semantics rely on several parameters. The only required value that needs to be set is either a Schedule-To-Close Timeout or a Start-To-Close Timeout. These values are set in the Activity Options.

Get Activity Execution results

How to get the results of an Activity Execution using the Temporal .NET SDK

The Activity result is the returned in the task from the ExecuteActivityAsync call.

Run Worker Process

How to create and run a Worker Process using the Temporal .NET SDK

The Worker Process is where Workflow Functions and Activity Functions are executed.

  • Each Worker Entity in the Worker Process must register the exact Workflow Types and Activity Types it may execute.
  • Each Worker Entity must also associate itself with exactly one Task Queue.
  • Each Worker Entity polling the same Task Queue must be registered with the same Workflow Types and Activity Types.

A Worker Entity is the component within a Worker Process that listens to a specific Task Queue.

Although multiple Worker Entities can be in a single Worker Process, a single Worker Entity Worker Process may be perfectly sufficient. For more information, see the Worker tuning guide.

A Worker Entity contains a Workflow Worker and/or an Activity Worker, which makes progress on Workflow Executions and Activity Executions, respectively.

To develop a Worker, create a new Temporalio.Worker.TemporalWorker providing the Client and worker options which include Task Queue, Workflows, and Activities and more. The following code example creates a Worker that polls for tasks from the Task Queue and executes the Workflow. When a Worker is created, it accepts a list of Workflows, a list of Activities, or both.

// Create a client to localhost on default namespace
var client = await TemporalClient.ConnectAsync(new("localhost:7233")
{
LoggerFactory = LoggerFactory.Create(builder =>
builder.
AddSimpleConsole(options => options.TimestampFormat = "[HH:mm:ss] ").
SetMinimumLevel(LogLevel.Information)),
});

// Cancellation token cancelled on ctrl+c
using var tokenSource = new CancellationTokenSource();
Console.CancelKeyPress += (_, eventArgs) =>
{
tokenSource.Cancel();
eventArgs.Cancel = true;
};

// Create an activity instance with some state
var activities = new MyActivities();

// Run worker until cancelled
Console.WriteLine("Running worker");
using var worker = new TemporalWorker(
client,
new TemporalWorkerOptions(taskQueue: "my-task-queue").
AddAllActivities(activities).
AddWorkflow<MyWorkflow>());
try
{
await worker.ExecuteAsync(tokenSource.Token);
}
catch (OperationCanceledException)
{
Console.WriteLine("Worker cancelled");
}

All Workers listening to the same Task Queue name must be registered to handle the exact same Workflows Types and Activity Types.

If a Worker polls a Task for a Workflow Type or Activity Type it does not know about, it fails that Task. However, the failure of the Task does not cause the associated Workflow Execution to fail.

Worker Processes with host builder and dependency injection

The Temporalio.Extensions.Hosting extension exists for .NET developers to support HostBuilder and Dependency Injection approaches.

To create the same worker as before using this approach:

var host = Host.CreateDefaultBuilder(args)
.ConfigureLogging(ctx => ctx.AddSimpleConsole().SetMinimumLevel(LogLevel.Information))
.ConfigureServices(ctx =>
ctx.
// Add the database client at the scoped level
AddScoped<IMyDatabaseClient, MyDatabaseClient>().
// Add the worker
AddHostedTemporalWorker(
clientTargetHost: "localhost:7233",
clientNamespace: "default",
taskQueue: "my-task-queue").
// Add the activities class at the scoped level
AddScopedActivities<MyActivities>().
AddWorkflow<MyWorkflow>())
.Build();
await host.RunAsync();

Set a Dynamic Workflow

How to set a Dynamic Workflow using the Temporal .NET SDK

A Dynamic Workflow in Temporal is a Workflow that is invoked dynamically at runtime if no other Workflow with the same name is registered. A Workflow can be made dynamic by setting Dynamic as true on the [Workflow] attribute. You must register the Workflow with the Worker before it can be invoked. Only one Dynamic Workflow can be present on a Worker.

The Workflow Definition must then accept a single argument of type Temporalio.Converters.IRawValue[]. The Workflow.PayloadConverter property is used to convert an IRawValue object to the desired type using extension methods in the Temporalio.Converters namespace.

[Workflow(Dynamic = true)]
public class DynamicWorkflow
{
[WorkflowRun]
public async Task<string> RunAsync(IRawValue[] args)
{
var name = Workflow.PayloadConverter.ToValue<string>(args.Single());
var param = MyActivityParams("Hello", name);
return await Workflow.ExecuteActivityAsync(
(MyActivities a) => a.MyActivity(param),
new() { StartToCloseTimeout = TimeSpan.FromMinutes(5) });
}
}

Set a Dynamic Activity

How to set a Dynamic Activity using the Temporal .NET SDK

A Dynamic Activity in Temporal is an Activity that is invoked dynamically at runtime if no other Activity with the same name is registered. An Activity can be made dynamic by setting Dynamic as true on the [Activity] attribute. You must register the Activity with the Worker before it can be invoked. Only one Dynamic Activity can be present on a Worker.

The Activity Definition must then accept a single argument of type Temporalio.Converters.IRawValue[]. The PayloadConverter property on the ActivityExecutionContext is used to convert an IRawValue object to the desired type using extension methods in the Temporalio.Converters namespace.

public class MyActivities
{
[Activity(Dynamic = true)]
public string DynamicActivity(IRawValue[] args)
{
var input = ActivityExecutionContext.Current.PayloadConverter.ToValue<MyActivityParams>(args.Single());
return $"{input.Greeting}, {input.Name}!";
}
}