.NET Generic Host

The Worker Service templates create a .NET Generic Host, HostBuilder. The Generic Host can be used with other types of .NET applications, such as Console apps.

A host is an object that encapsulates an app's resources and lifetime functionality, such as:

  • Dependency injection (DI)
  • Logging
  • Configuration
  • App shutdown
  • IHostedService implementations

When a host starts, it calls IHostedService.StartAsync on each implementation of IHostedService registered in the service container's collection of hosted services. In a worker service app, all IHostedService implementations that contain BackgroundService instances have their BackgroundService.ExecuteAsync methods called.

The main reason for including all of the app's interdependent resources in one object is lifetime management: control over app startup and graceful shutdown. Reference the Microsoft.Extensions.Hosting NuGet package to achieve lifetime management.

Set up a host

The host is typically configured, built, and run by code in the Program class. The Main method:

The .NET Worker Service templates generate the following code to create a Generic Host:

using Example.WorkerService;

HostApplicationBuilder builder = Host.CreateApplicationBuilder(args);

IHost host = builder.Build();

Default builder settings

The CreateApplicationBuilder method:

  • Sets the content root to the path returned by GetCurrentDirectory().
  • Loads host configuration from:
    • Environment variables prefixed with DOTNET_.
    • Command-line arguments.
  • Loads app configuration from:
    • appsettings.json.
    • appsettings.{Environment}.json.
    • Secret Manager when the app runs in the Development environment.
    • Environment variables.
    • Command-line arguments.
  • Adds the following logging providers:
    • Console
    • Debug
    • EventSource
    • EventLog (only when running on Windows)
  • Enables scope validation and dependency validation when the environment is Development.

The HostApplicationBuilder.Services is an Microsoft.Extensions.DependencyInjection.IServiceCollection instance. These services are used to build an IServiceProvider that's used with dependency injection to resolve the registered services.

Framework-provided services

The following services are registered automatically:


Inject the IHostApplicationLifetime service into any class to handle post-startup and graceful shutdown tasks. Three properties on the interface are cancellation tokens used to register app start and app stop event handler methods. The interface also includes a StopApplication() method.

The following example is an IHostedService implementation that registers IHostApplicationLifetime events:

using Microsoft.Extensions.Hosting;
using Microsoft.Extensions.Logging;

namespace AppLifetime.Example;

public sealed class ExampleHostedService : IHostedService
    private readonly ILogger _logger;

    public ExampleHostedService(
        ILogger<ExampleHostedService> logger,
        IHostApplicationLifetime appLifetime)
        _logger = logger;


    public Task StartAsync(CancellationToken cancellationToken)
        _logger.LogInformation("1. StartAsync has been called.");

        return Task.CompletedTask;

    public Task StopAsync(CancellationToken cancellationToken)
        _logger.LogInformation("4. StopAsync has been called.");

        return Task.CompletedTask;

    private void OnStarted()
        _logger.LogInformation("2. OnStarted has been called.");

    private void OnStopping()
        _logger.LogInformation("3. OnStopping has been called.");

    private void OnStopped()
        _logger.LogInformation("5. OnStopped has been called.");

The Worker Service template could be modified to add the ExampleHostedService implementation:

using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Hosting;
using AppLifetime.Example;

HostApplicationBuilder builder = Host.CreateApplicationBuilder(args);

using IHost host = builder.Build();

await host.RunAsync();

The application would write the following sample output:

// Sample output:
//     info: ExampleHostedService[0]
//           1. StartAsync has been called.
//     info: ExampleHostedService[0]
//           2. OnStarted has been called.
//     info: Microsoft.Hosting.Lifetime[0]
//           Application started.Press Ctrl+C to shut down.
//     info: Microsoft.Hosting.Lifetime[0]
//           Hosting environment: Production
//     info: Microsoft.Hosting.Lifetime[0]
//           Content root path: ..\app-lifetime\bin\Debug\net7.0
//     info: ExampleHostedService[0]
//           3. OnStopping has been called.
//     info: Microsoft.Hosting.Lifetime[0]
//           Application is shutting down...
//     info: ExampleHostedService[0]
//           4. StopAsync has been called.
//     info: ExampleHostedService[0]
//           5. OnStopped has been called.


The IHostLifetime implementation controls when the host starts and when it stops. The last implementation registered is used. Microsoft.Extensions.Hosting.Internal.ConsoleLifetime is the default IHostLifetime implementation. For more information on the lifetime mechanics of shutdown, see Host shutdown.


Inject the IHostEnvironment service into a class to get information about the following settings:

Host configuration

Host configuration is used to configure properties of the IHostEnvironment implementation.

The host configuration is available in HostBuilderContext.Configuration within the ConfigureAppConfiguration method. When you call the ConfigureAppConfiguration method, the HostBuilderContext and IConfigurationBuilder are passed into the configureDelegate. The configureDelegate is defined as an Action<HostBuilderContext, IConfigurationBuilder>. The host builder context exposes the Configuration property, which is an instance of IConfiguration. It represents the configuration built from the host, whereas the IConfigurationBuilder is the builder object used to configure the app.


After ConfigureAppConfiguration is called the HostBuilderContext.Configuration is replaced with the app config.

To add host configuration, call ConfigureHostConfiguration on IHostBuilder. ConfigureHostConfiguration can be called multiple times with additive results. The host uses whichever option sets a value last on a given key.

The following example creates host configuration:

using Microsoft.Extensions.Configuration;
using Microsoft.Extensions.Hosting;

HostApplicationBuilder builder = Host.CreateApplicationBuilder(args);

builder.Environment.ContentRootPath = Directory.GetCurrentDirectory();
builder.Configuration.AddJsonFile("hostsettings.json", optional: true);
builder.Configuration.AddEnvironmentVariables(prefix: "PREFIX_");

using IHost host = builder.Build();

// Application code should start here.

await host.RunAsync();

App configuration

App configuration is created by calling ConfigureAppConfiguration on IHostBuilder. ConfigureAppConfiguration can be called multiple times with additive results. The app uses whichever option sets a value last on a given key.

The configuration created by ConfigureAppConfiguration is available in HostBuilderContext.Configuration for subsequent operations and as a service from DI. The host configuration is also added to the app configuration.

For more information, see Configuration in .NET.

Host shutdown

There are several ways in which a hosted process is stopped. Most commonly, a hosted process can be stopped in the following ways:

The hosting code isn't responsible for handling these scenarios. The owner of the process needs to deal with them the same as any other app. There are several other ways in which a hosted service process can be stopped:

  • If ConsoleLifetime is used (UseConsoleLifetime), it listens for the following signals and attempts to stop the host gracefully.
    • SIGINT (or CTRL+C).
    • SIGQUIT (or CTRL+BREAK on Windows, CTRL+\ on Unix).
    • SIGTERM (sent by other apps, such as docker stop).
  • If the app calls Environment.Exit.

The built-in hosting logic handles these scenarios, specifically the ConsoleLifetime class. ConsoleLifetime tries to handle the "shutdown" signals SIGINT, SIGQUIT, and SIGTERM to allow for a graceful exit to the application.

Before .NET 6, there wasn't a way for .NET code to gracefully handle SIGTERM. To work around this limitation, ConsoleLifetime would subscribe to System.AppDomain.ProcessExit. When ProcessExit was raised, ConsoleLifetime would signal the host to stop and block the ProcessExit thread, waiting for the host to stop.

The process exit handler would allow for the clean-up code in the application to run—for example, IHost.StopAsync and code after HostingAbstractionsHostExtensions.Run in the Main method.

However, there were other issues with this approach because SIGTERM wasn't the only way ProcessExit was raised. SIGTERM is also raised when app code calls Environment.Exit. Environment.Exit isn't a graceful way of shutting down a process in the Microsoft.Extensions.Hosting app model. It raises the ProcessExit event and then exits the process. The end of the Main method doesn't get executed. Background and foreground threads are terminated, and finally blocks aren't executed.

Since ConsoleLifetime blocked ProcessExit while waiting for the host to shut down, this behavior led to deadlocks from Environment.Exit also blocks waiting for the call to ProcessExit. Additionally, since the SIGTERM handling was attempting to gracefully shut down the process, ConsoleLifetime would set the ExitCode to 0, which clobbered the user's exit code passed to Environment.Exit.

In .NET 6, POSIX signals are supported and handled. The ConsoleLifetime handles SIGTERM gracefully, and no longer gets involved when Environment.Exit is invoked.


For .NET 6+, ConsoleLifetime no longer has logic to handle scenario Environment.Exit. Apps that call Environment.Exit and need to perform clean-up logic can subscribe to ProcessExit themselves. Hosting will no longer attempt to gracefully stop the host in these scenarios.

If your application uses hosting, and you want to gracefully stop the host, you can call IHostApplicationLifetime.StopApplication instead of Environment.Exit.

Hosting shutdown process

The following sequence diagram shows how the signals are handled internally in the hosting code. Most users don't need to understand this process. But for developers that need a deep understanding, a good visual may help you get started.

After the host has been started, when a user calls Run or WaitForShutdown, a handler gets registered for IApplicationLifetime.ApplicationStopping. Execution is paused in WaitForShutdown, waiting for the ApplicationStopping event to be raised. The Main method doesn't return right away, and the app stays running until Run or WaitForShutdown returns.

When a signal is sent to the process, it initiates the following sequence:

Hosting shutdown sequence diagram.

  1. The control flows from ConsoleLifetime to ApplicationLifetime to raise the ApplicationStopping event. This signals WaitForShutdownAsync to unblock the Main execution code. In the meantime, the POSIX signal handler returns with Cancel = true since the POSIX signal has been handled.
  2. The Main execution code starts executing again and tells the host to StopAsync(), which in turn stops all the hosted services, and raises any other stopped events.
  3. Finally, WaitForShutdown exits, allowing for any application clean-up code to execute, and for the Main method to exit gracefully.

Host shutdown in web server scenarios

There are various other common scenarios in which graceful shutdown works in Kestrel for both HTTP/1.1 and HTTP/2 protocols, and how you can configure it in different environments with a load balancer to drain traffic smoothly. While web server configuration is beyond the scope of this article, you can find more information on Configure options for the ASP.NET Core Kestrel web server documentation.

When the Host receives a shutdown signal (for example, CTL+C or StopAsync), it notifies the application by signaling ApplicationStopping. You should subscribe to this event if you have any long-running operations that need to finish gracefully.

Next, the Host calls IServer.StopAsync with a shutdown timeout that you can configure (default 30s). Kestrel (and Http.Sys) close their port bindings and stop accepting new connections. They also tell the current connections to stop processing new requests. For HTTP/2 and HTTP/3, a preliminary GOAWAY message is sent to the client. For HTTP/1.1, they stop the connection loop because requests are processed in order. IIS behaves differently, by rejecting new requests with a 503 status code.

The active requests have until the shutdown timeout to complete. If they're all complete before the timeout, the server returns control to the host sooner. If the timeout expires, the pending connections and requests are aborted forcefully, which can cause errors in the logs and to the clients.

Load balancer considerations

To ensure a smooth transition of clients to a new destination when working with a load balancer, you can follow these steps:

  • Bring up the new instance and start balancing traffic to it (you may already have several instances for scaling purposes).
  • Disable or remove the old instance in the load balancer configuration so it stops receiving new traffic.
  • Signal the old instance to shut down.
  • Wait for it to drain or time out.

See also