I have written a Windows Service project which hosts a long-running message pump task which is meant to run for the duration of the service. When the service starts, it starts the task. When the service stops, it stops the task:
void OnStart()
{
MessagePump.Start();
}
void OnStop()
{
MessagePump.Stop();
}
Where MessagePump.Start does a Task.Factory.StartNew, and MessagePump.Stop signals the task to stop and does a Wait().
So far so good, but I'm wondering how best to handle faults. If the task has an unhandled exception, I'd want the service to stop but since nothing is typically Wait-ing on the task, I imagine it'll just sit doing nothing. How can I elegantly handle this situation?
UPDATE:
The consensus seems to be using 'await' or ContinueWith. Here is how I'm currently coding my Start method to use this:
public async static void Start()
{
this.state = MessagePumpState.Running;
this.task = Task.Factory.StartNew(() => this.ProcessLoop(), TaskCreationOptions.LongRunning);
try
{
await this.task;
}
catch
{
this.state = MessagePumpState.Faulted;
throw;
}
}
Make you MessagePump.Start() method return the task. Then
MessagePump.Start().ContinueWith(t =>
{
// handle exception
},
TaskContinuationOptions.OnlyOnFaulted);
UPDATE:
I would do the next:
private MessagePump _messagePump;
async void OnStart()
{
this._messagePump = new MessagePump();
try
{
// make Start method return the task to be able to handle bubbling exception here
await _messagePump.Start();
}
catch (Exception ex)
{
// log exception
// abort service
}
}
void OnStop()
{
_messagePump.Stop();
}
public enum MessagePumpState
{
Running,
Faulted
}
public class MessagePump
{
private CancellationTokenSource _cancallationTokenSrc;
private MessagePumpState _state;
public async Task Start()
{
if (_cancallationTokenSrc != null)
{
throw new InvalidOperationException("Task is already running!");
}
this._state = MessagePumpState.Running;
_cancallationTokenSrc = new CancellationTokenSource();
var task = Task.Factory.StartNew(() => this.ProcessLoop(_cancallationTokenSrc.Token), _cancallationTokenSrc.Token);
try
{
await task;
}
catch
{
this._state = MessagePumpState.Faulted;
throw;
}
}
public void Stop()
{
if (_cancallationTokenSrc != null)
{
_cancallationTokenSrc.Cancel();
_cancallationTokenSrc = null;
}
}
public void ProcessLoop(CancellationToken token)
{
// check if task has been canceled
while (!token.IsCancellationRequested)
{
Console.WriteLine(DateTime.Now);
Thread.Sleep(1000);
}
}
}
You can try something like this :
void OnStart()
{
MessagePump.StartAsync();
MessagePump.ErrorEvent += OnError();
}
Then your StartAsync will look something like:
public async Task StartAsync()
{
// your process
// if error, send event to messagePump
}
And if you decide to use Tasks, then it is better to use Task.Run and not Task.Factory.StartNew()
Related
I don't think this question is a duplicate of "Proper way to deal with exceptions in DisposeAsync".
Let's say my class that implements IAsynsDisposable because it has a long-running background task, and DisposeAsync terminates that task. A familiar pattern might be the Completion property, e.g. ChannelReader<T>.Completion (despite ChannelReader doesn't implement IAsynsDisposable).
Is it considered a good practice to propagate the Completion task's exceptions outside DisposeAsync?
Here is a complete example that can be copied/pasted into a dotnet new console project. Note await this.Completion inside DisposeAsync:
try
{
await using var service = new BackgroundService(TimeSpan.FromSeconds(2));
await Task.Delay(TimeSpan.FromSeconds(3));
}
catch (Exception ex)
{
Console.WriteLine(ex);
Console.ReadLine();
}
class BackgroundService: IAsyncDisposable
{
public Task Completion { get; }
private CancellationTokenSource _diposalCts = new();
public BackgroundService(TimeSpan timeSpan)
{
this.Completion = Run(timeSpan);
}
public async ValueTask DisposeAsync()
{
_diposalCts.Cancel();
try
{
await this.Completion;
}
finally
{
_diposalCts.Dispose();
}
}
private async Task Run(TimeSpan timeSpan)
{
try
{
await Task.Delay(timeSpan, _diposalCts.Token);
throw new InvalidOperationException("Boo!");
}
catch (OperationCanceledException)
{
}
}
}
Alternatively, I can observe service.Completion explicitly in the client code (and ignore its exceptions inside DiposeAsync to avoid them being potentially thrown twice), like below:
try
{
await using var service = new BackgroundService(TimeSpan.FromSeconds(2));
await Task.Delay(TimeSpan.FromSeconds(3));
await service.Completion;
}
catch (Exception ex)
{
Console.WriteLine(ex);
Console.ReadLine();
}
class BackgroundService: IAsyncDisposable
{
public Task Completion { get; }
private CancellationTokenSource _diposalCts = new();
public BackgroundService(TimeSpan timeSpan)
{
this.Completion = Run(timeSpan);
}
public async ValueTask DisposeAsync()
{
_diposalCts.Cancel();
try
{
await this.Completion;
}
catch
{
// the client should observe this.Completion
}
finally
{
_diposalCts.Dispose();
}
}
private async Task Run(TimeSpan timeSpan)
{
try
{
await Task.Delay(timeSpan, _diposalCts.Token);
throw new InvalidOperationException("Boo!");
}
catch (OperationCanceledException)
{
}
}
}
Is there a concensus about which option is better?
For now, I've settled on a reusable helper class LongRunningAsyncDisposable (here's a gist, warning: barely tested yet), which allows:
to start a background task;
stop this task (via a cancellation token) by calling IAsyncDisposable.DisposeAsync at any time, in a thread-safe, concurrency-friendly way;
configure whether DisposeAsync should re-throw the task's exceptions (DisposeAsync will await the task's completion either way, before doing a cleanup);
observe the task's status, result and exceptions at any time via LongRunningAsyncDisposable.Completion property.
I have a code block which is eventually accessed by multiple threads. I search for an up to date async mechanism to continue executing when all threads have passed.
Currently I do the following with a CountDownEvent which works just fine (without async support).
public class Watcher
{
private static readonly Logger Log = LogManager.GetCurrentClassLogger();
private readonly CountdownEvent _isUpdating = new CountdownEvent(1);
private readonly IActivity _activity;
public Watcher([NotNull] IActivity activity)
{
_activity = activity ?? throw new ArgumentNullException(nameof(activity));
_activity.Received += OnReceived;
}
private void OnReceived(IReadOnlyCollection<Summary> summaries)
{
_isUpdating.AddCount();
try
{
// Threads processing
}
finally
{
_isUpdating.Signal();
}
}
private void Disable()
{
_activity.Received -= OnReceived;
_isUpdating.Signal();
/* await */ _isUpdating.Wait();
}
}
Do I need to use any of those AsyncCountdownEvent implementations or is there any other built-in mechanism? I already thought about using a BufferBlock because it has async functionality but I think it's a bit overkill.
Additional to the comments:
IActivity is a WebService call (but shouldn't effect the implementation on top or vice versa)
public async Task Start(bool alwayRetry = true, CancellationToken cancellationToken = new CancellationToken())
{
var milliseconds = ReloadSeconds * 1000;
do
{
try
{
var summaries = await PublicAPI.GetSummariesAsync(cancellationToken).ConfigureAwait(false);
OnSummariesReceived(summaries);
}
catch (Exception ex)
{
Log.Error(ex.Message);
OnErrorOccurred(ex);
}
await Task.Delay(milliseconds, cancellationToken).ConfigureAwait(false);
// ReSharper disable once LoopVariableIsNeverChangedInsideLoop
} while (alwayRetry);
}
It's not clear the IActivity signatures; but you can wait for a range of tasks to be completed:
class MultiAsyncTest {
Task SomeAsync1() { return Task.Delay(1000); }
Task SomeAsync2() { return Task.Delay(2000);}
Task EntryPointAsync() {
var tasks = new List<Task>();
tasks.Add(SomeAsync1());
tasks.Add(SomeAsync2());
return Task.WhenAll(tasks);
}
}
What's IActivity's signature? Does it support Task? Or you are using Thread? More explanation would help to a more specified answer.
It seems like right after I call my first async method (GetBar() in this example), the CancellationToken's IsCancellationRequested is set to true, but I don't want that and don't understand why it's happening.
This is in an Azure Cloud Service worker role, if that matters.
public class WorkerRole : RoleEntryPoint
{
private CancellationTokenSource cancellationTokenSource;
private Task runTask;
public override void Run()
{
this.cancellationTokenSource = new CancellationTokenSource();
this.runTask = Task.Run(() => Foo.Bar(this.cancellationTokenSource.Token), this.cancellationTokenSource.Token);
}
public override void OnStop()
{
this.cancellationTokenSource.Cancel();
try
{
this.runTask.Wait();
}
catch (Exception e)
{
Logger.Error(e, e.Message);
}
base.OnStop();
}
// ... OnStart omitted
}
public static class Foo
{
public static async Bar(CancellationToken token)
{
while (true)
{
try
{
token.ThrowIfCancellationRequested();
var bar = await FooService.GetBar().ConfigureAwait(false);
// Now token.IsCancellationRequested == true. Why? The above call does not take the token as input.
}
catch (OperationCanceledException)
{
// ... Handling
}
}
}
}
I've successfully used CancellationTokens once before in another project and I use a similar setup here. The only difference I'm aware of is that this is in an Azure Cloud Service. Any idea why IsCancellationRequested is getting set to true?
It appears OnStop was called while you where awaiting for FooService.GetBar() to complete. Perhaps add some form of logging to see if OnStop is called between the token.ThrowIfCancellationRequested(); and after the var bar = await ... returns to confirm.
That is what is causing the token to be canceled.
To solve the problem you need to make sure the overridden Run method does not return till the work is complete.
public override void Run()
{
this.cancellationTokenSource = new CancellationTokenSource();
this.runTask = Task.Run(() => Foo.Bar(this.cancellationTokenSource.Token), this.cancellationTokenSource.Token);
this.runTask.Wait(); //You may need a try/catch around it
}
I have a situation where I must call an async method synchronously, and it is done so as follows:
obj.asyncMethod().Wait(myCancelToken)
If the cancellation token is switched the disposable's within the task will not get disposed despite being activated via a using statement.
The below program illustrates the problem:
using System;
using System.Threading;
using System.Threading.Tasks;
namespace LeakTest {
class Program {
static void Main(string[] args) {
try {
var timeout = new CancellationTokenSource(TimeSpan.FromMilliseconds(100));
LongRunningTask().Wait(timeout.Token);
} catch (OperationCanceledException error) {
// handling timeout is logically okay, but expect nothing to be leaked
}
Console.WriteLine("Leaked Instances = {0}", DisposableResource.Instances);
Console.ReadKey();
}
static async Task LongRunningTask() {
using (var resource = new DisposableResource()) {
await Task.Run( () => Thread.Sleep(1000));
}
}
public class DisposableResource : IDisposable {
public static int Instances = 0;
public DisposableResource() {
Instances++;
}
public void Dispose() {
Instances--;
}
}
}
}
It seems Wait method just kills the task thread on cancellation instead of triggering an exception within that thread and letting it terminate naturally. Question is why?
You've cancelled the task returned by Wait(timeout.Token) not the one returned from LongRunningTask, if you want to cancel that one pass the token to Task.Run and also use await Task.Delay instead of Thread.Sleep and pass the token there as well.
static void Main(string[] args)
{
try
{
var timeout = new CancellationTokenSource(TimeSpan.FromMilliseconds(100));
LongRunningTask(timeout.Token).Wait();
}
catch (AggregateException error)
{
// handling timeout is logically okay, but expect nothing to be leaked
}
Console.WriteLine("Leaked Instances = {0}", DisposableResource.Instances);
Console.ReadLine();
}
static async Task LongRunningTask(CancellationToken token)
{
using (var resource = new DisposableResource())
{
await Task.Run(async () => await Task.Delay(1000, token), token);
}
}
public class DisposableResource : IDisposable
{
public static int Instances = 0;
public DisposableResource()
{
Instances++;
}
public void Dispose()
{
Instances--;
}
}
Note that the using statment will still dispose of the resource once the long running operation finishes. Run this example:
static void Main(string[] args)
{
try {
var timeout = new CancellationTokenSource(TimeSpan.FromMilliseconds(100));
LongRunningTask().Wait(timeout.Token);
} catch (OperationCanceledException error) {
// handling timeout is logically okay, but expect nothing to be leaked
}
Console.WriteLine("Leaked Instances = {0}", DisposableResource.Instances);
Console.ReadKey();
}
static async Task LongRunningTask()
{
using (var resource = new DisposableResource())
{
await Task.Run(() => Thread.Sleep(1000));
}
}
public class DisposableResource : IDisposable
{
public static int Instances = 0;
public DisposableResource()
{
Instances++;
}
public void Dispose()
{
Instances--;
Console.WriteLine("Disposed resource. Leaked Instances = {0}", Instances);
}
}
Output
Leaked Instances = 1
Disposed resource. Leaked Instances = 0
It seems Wait method just kills the task thread on cancellation instead of triggering an exception within that thread
You are incorrect, on when you cancel the only thing that happens is you stop waiting for Wait(myCancelToken) to complete, the task is still running in the background.
In order to cancel the background task you must pass the cancelation token into all of the methods down the chain. If you want the innermost layer (the long running one) to stop early that code must call token.ThrowIfCancellationRequested() throughout its code.
This example "fails":
static async void Main(string[] args)
{
try
{
await TaskEx.Run(() => { throw new Exception("failure"); });
}
catch (Exception)
{
throw new Exception("success");
}
}
That is, the exception with the text "failure" bubbles up.
Then I tried this workaround:
static async void Main(string[] args)
{
try
{
await SafeRun(() => { throw new Exception("failure"); });
}
catch (Exception)
{
throw new Exception("success");
}
}
static async Task SafeRun(Action action)
{
var ex = default(Exception);
await TaskEx.Run(() =>
{
try
{
action();
}
catch (Exception _)
{
ex = _;
}
});
if (ex != default(Exception))
throw ex;
}
That didn't help either.
I suppose my Async CTP refresh installation could be hosed.
Should this code work as I expect ("success" bubbles up, not "failure"), or is this not "supposed" to work that way. And if not, how would you work around it?
The behavior you are seeing is likely an edge case bug or may even be correct, if unintuitive. Normally when you invoke an async method synchronously, it wraps a task around to execute and since there is no one waiting on the task to finish, the exception never makes it to the main thread. If you were to call Main directly it would succeed, but then your runtime would see an exception of "success" on another thread.
Since main is the entrypoint of your application, it is invoked synchronously and likely as the entrypoint doesn't trigger the Task wrapping behavior, so that await isn't run properly and the TaskEx.Run throws on its own thread, which shows up in the runtime as an exception being thrown on another thread.
If you were to run main as an async method, i.e. returning a Task (since an async that returns void can only really be called via await) and blocking on it from your synchronous main context, you would get the appropriate behavior as the below test illustrates:
static async Task Main() {
try {
await TaskEx.Run(() => { throw new Exception("failure"); });
} catch(Exception) {
throw new Exception("success");
}
}
static async Task Main2() {
await Main();
}
[Test]
public void CallViaAwait() {
var t = Main2();
try {
t.Wait();
Assert.Fail("didn't throw");
} catch(AggregateException e) {
Assert.AreEqual("success",e.InnerException.Message);
}
}
[Test]
public void CallDirectly() {
var t = Main();
try {
t.Wait();
Assert.Fail("didn't throw");
} catch(AggregateException e) {
Assert.AreEqual("success", e.InnerException.Message);
}
}
I.e. the Task faults with an AggregateException which contains the success exception as it's inner exception.