Multiple threads awaiting Task if it is already running - c#

I have app which have different timers running in the background doing synchronization with the API.
However, before the actual sync they all ping the API whether they can proceed or not. In order to not make multiple calls, I've made them await the same Task if it is already running. After the result comes from the task those who await should get the result and in the mean time if there is a call to the CanExecute method and the Task is not running it should rerun the Task and so on.
protected async Task<bool> CanExecute([CallerMemberName] string memberName = "")
{
lock (_lockobj)
{
if (_runningTask == null)
{
_runningTask = CanExecuteTask(memberName);
}
}
var result = await _runningTask;
_runningTask = null;
return result;
}
private async Task<bool> CanExecuteTask(string callingMemberName)
{
var result = // do http call and some other method calls
return result;
}
I'm not sure what the problem is, but I think it happens to deadlock and the synchronization doesn't proceed. What is the proper way to achieve that?

As written this code is not safe - you both read and write _runningTask outside of the lock. So for example, one thread can execute
_runningTask = null;
While another thread just got past the lock statement while task was not null and is about to execute await _runningTask. This will result in await null so in NullReferenceException.
You can avoid that while keeping the same principle by doing this:
private Task<bool> _runningTask = Task.FromResult(false);
protected async Task<bool> CanExecute([CallerMemberName] string memberName = "") {
Task<bool> task;
lock (_lockobj) {
if (_runningTask.IsCompleted) {
_runningTask = CanExecuteTask(memberName);
}
task = _runningTask;
}
var result = await task;
return result;
}
Instead of setting task to null and then checking that - we check if task has been completed (which means succeeded, faulted, or cancelled). We then also start with already completed task instead of null.
Also we assign _runningTask to local variable inside the lock, and then await that, to avoid reading _runningTask field outside of the lock.
Awaiting the same task from multiple threads is safe, so we should be fine here.

Related

What task is being returned and why is this task status RanToCompletion

I am rather new to task based programming and trying to determine how to return a task and verify that it has been started. The code that I got to work was not what I was expecting. The console application is as follows:
public static void Main(string[] args)
{
var mySimple = new Simple();
var cts = new CancellationTokenSource();
var task = mySimple.RunSomethingAsync(cts.Token);
while (task.Status != TaskStatus.RanToCompletion)
{
Console.WriteLine("Starting...");
Thread.Sleep(100);
}
Console.WriteLine("It is started");
Console.ReadKey();
cts.Cancel();
}
public class Simple
{
public async void RunSomething(CancellationToken token)
{
var count = 0;
while (true)
{
if (token.IsCancellationRequested)
{
break;
}
Console.WriteLine(count++);
await Task.Delay(TimeSpan.FromMilliseconds(1000), token).ContinueWith(task => { });
}
}
public Task RunSomethingAsync(CancellationToken token)
{
return Task.Run(() => this.RunSomething(token));
}
}
The output is:
Starting...
0
It is started
1
2
3
4
Why is the task that is being returned have a status as TaskStatus.RanToCompletion compared to TaskStatus.Running as we see that the while loop is still executing? Am I checking the status of the task of putting the RunSomething task on the threadpool rather than the RunSomething task itself?
RunSomething is an async void method, meaning it exposes no means of the caller ever determining when it finishes, they can only ever start the operation and then have no idea what happens next. You then wrap a call to it inside of Task.Run, this is schedluing a thread pool thread to start RunSomething. It will then complete as soon as it has finished starting that Task.
If RunSomething actually returned a Task, then the caller would be able to determine when it actually finished, and if you waited on it it wouldn't actually indicate that it was done until that asynchronous operation was actually finished (there would be no reason to use Task.Run to start it in another thead, you'd be better off just calling it directly and not wasting the effort of moving that to a thread pool thread).
Never use async void (https://msdn.microsoft.com/en-us/magazine/jj991977.aspx)
instead you should use async Task
If you need to call an async method from a non-async (such as from a static void main) you should do something like this:
mySimple.RunSomethingAsync(cts.Token).GetAwaiter().GetResult();
That will effectively make the method a synchronous call.
You can use async void, but only for events.

Task being marked as RanToCompletion at await, when still Running

I'm still getting up to speed with async & multi threading. I'm trying to monitor when the Task I Start is still running (to show in a UI). However it's indicating that it is RanToCompletion earlier than I want, when it hits an await, even when I consider its Status as still Running.
Here is the sample I'm doing. It all seems to be centred around the await's. When it hits an await, it is then marked as RanToCompletion.
I want to keep track of the main Task which starts it all, in a way which indicates to me that it is still running all the way to the end and only RanToCompletion when it is all done, including the repo call and the WhenAll.
How can I change this to get the feedback I want about the tskProdSeeding task status?
My Console application Main method calls this:
Task tskProdSeeding;
tskProdSeeding = Task.Factory.StartNew(SeedingProd, _cts.Token);
Which the runs this:
private async void SeedingProd(object state)
{
var token = (CancellationToken)state;
while (!token.IsCancellationRequested)
{
int totalSeeded = 0;
var codesToSeed = await _myRepository.All().ToListAsync(token);
await Task.WhenAll(Task.Run(async () =>
{
foreach (var code in codesToSeed)
{
if (!token.IsCancellationRequested)
{
try
{
int seedCountByCode = await _myManager.SeedDataFromLive(code);
totalSeeded += seedCountByCode;
}
catch (Exception ex)
{
_logger.InfoFormat(ex.ToString());
}
}
}
}, token));
Thread.Sleep(30000);
}
}
If you use async void the outer task can't tell when the task is finished, you need to use async Task instead.
Second, once you do switch to async Task, Task.Factory.StartNew can't handle functions that return a Task, you need to switch to Task.Run(
tskProdSeeding = Task.Run(() => SeedingProd(_cts.Token), _cts.Token);
Once you do both of those changes you will be able to await or do a .Wait() on tskProdSeeding and it will properly wait till all the work is done before continuing.
Please read "Async/Await - Best Practices in Asynchronous Programming" to learn more about not doing async void.
Please read "StartNew is Dangerous" to learn more about why you should not be using StartNew the way you are using it.
P.S. In SeedingProd you should switch it to use await Task.Delay(30000); insetad of Thread.Sleep(30000);, you will then not tie up a thread while it waits. If you do this you likely could drop the
tskProdSeeding = Task.Run(() => SeedingProd(_cts.Token), _cts.Token);
and just make it
tskProdSeeding = SeedingProd(_cts.Token);
because the function no-longer has a blocking call inside of it.
I'm not convinced that you need a second thread (Task.Run or StartNew) at all. It looks like the bulk of the work is I/O-bound and if you're doing it asynchronously and using Task.Delay instead of Thread.Sleep, then there is no thread consumed by those operations and your UI shouldn't freeze. The first thing anyone new to async needs to understand is that it's not the same thing as multithreading. The latter is all about consuming more threads, the former is all about consuming fewer. Focus on eliminating the blocking and you shouldn't need a second thread.
As others have noted, SeedingProd needs to return a Task, not void, so you can observe its completion. I believe your method can be reduced to this:
private async Task SeedingProd(CancellationToken token)
{
while (!token.IsCancellationRequested)
{
int totalSeeded = 0;
var codesToSeed = await _myRepository.All().ToListAsync(token);
foreach (var code in codesToSeed)
{
if (token.IsCancellationRequested)
return;
try
{
int seedCountByCode = await _myManager.SeedDataFromLive(code);
totalSeeded += seedCountByCode;
}
catch (Exception ex)
{
_logger.InfoFormat(ex.ToString());
}
}
await Task.Dealy(30000);
}
}
Then simply call the method, without awaiting it, and you'll have your task.
Task mainTask = SeedingProd(token);
When you specify async on a method, it compiles into a state machine with a Task, so SeedingProd does not run synchronously, but acts as a Task even if returns void. So when you call Task.Factory.StartNew(SeedingProd) you start a task that kick off another task - that's why the first one finishes immediately before the second one. All you have to do is add the Task return parameter instead of void:
private async Task SeedingProdAsync(CancellationToken ct)
{
...
}
and call it as simply as this:
Task tskProdSeeding = SeedingProdAsync(_cts.Token);

Execute all tasks simultaneously and await their completion?

I have a series of async methods that I would like to execute simultaneously. Each of these methods return true or false in regards to if they execute successfully or not. Their results are also logged in our audit trail so that we can diagnose issues.
Some of my functions are not dependent on all of these methods executing successfully, and we fully expect some of them to fail from time to time. If they do fail, the program will continue to execute and it will merely alert our support staff that they need to correct the issue.
I'm trying to figure out what would be the best method for all of these functions to execute simultaneously, and yet have the parent function await them only after they have all begun to execute. The parent function will return False if ANY of the functions fail, and this will alert my application to cease execution.
My idea was to do something similar to:
public async Task<bool> SetupAccessControl(int objectTypeId, int objectId, int? organizationId)
{
using (var context = new SupportContext(CustomerId))
{
if (organizationId == null)
{
var defaultOrganization = context.Organizations.FirstOrDefault(n => n.Default);
if (defaultOrganization != null) organizationId = defaultOrganization.Id;
}
}
var acLink = AcLinkObjectToOrganiation(organizationId??0,objectTypeId,objectId);
var eAdmin = GrantRoleAccessToObject("eAdmin", objectTypeId, objectId);
var defaultRole = GrantRoleAccessToObject("Default", objectTypeId, objectId);
await acLink;
await eAdmin;
await defaultRole;
var userAccess = (objectTypeId != (int)ObjectType.User) || await SetupUserAccessControl(objectId);
return acLink.Result && eAdmin.Result && defaultRole.Result && userAccess;
}
public async Task<bool> SetupUserAccessControl(int objectId)
{
var everyoneRole = AddToUserRoleBridge("Everyone", objectId);
var defaultRole = AddToUserRoleBridge("Default", objectId);
await everyoneRole;
await defaultRole;
return everyoneRole.Result && defaultRole.Result;
}
Is there a better option? Should I restructure in any way? I'm simply trying to speed up execution time as I have a parent function that executes close to 20 other functions that are all independent of each other. Even at it's slowest, without async, it only takes about 1-2 seconds to execute. However, this will be scaled out to eventually have that parent call executed several hundred times (bulk insertions).
Async methods have a synchronous part which is the part before the first await of an uncompleted task is reached (if there isn't one then the whole method runs synchronously). That part is executed synchronously using the calling thread.
If you want to run these methods concurrently without parallelizing these parts simply invoke the methods, gather the tasks and use Task.WhenAll to await for all of them at once. When all tasks completed you can check the individual results:
async Task<bool> SetupUserAccessControlAsync(int objectId)
{
var everyoneRoleTask = AddToUserRoleBridgeAsync("Everyone", objectId);
var defaultRoleTask = AddToUserRoleBridgeAsync("Default", objectId);
await Task.WhenAll(everyoneRoleTask, defaultRoleTask)
return await everyoneRoleTask && await defaultRoleTask;
}
If you do want to parallelize that synchronous part as well you need multiple threads so instead of simply invoking the async method, use Task.Run to offload to a ThreadPool thread:
async Task<bool> SetupUserAccessControlAsync(int objectId)
{
var everyoneRoleTask = Task.Run(() => AddToUserRoleBridgeAsync("Everyone", objectId));
var defaultRoleTask = Task.Run(() => AddToUserRoleBridgeAsync("Default", objectId));
await Task.WhenAll(everyoneRoleTask, defaultRoleTask)
return await everyoneRoleTask && await defaultRoleTask;
}
If all your methods return bool you can gather all the tasks in a list, get the results from Task.WhenAll and check whether all returned true:
async Task<bool> SetupUserAccessControlAsync(int objectId)
{
var tasks = new List<Task<bool>>();
tasks.Add(AddToUserRoleBridgeAsync("Everyone", objectId));
tasks.Add(AddToUserRoleBridgeAsync("Default", objectId));
return (await Task.WhenAll(tasks)).All(_ => _);
}

Regarding the usage of SemaphoreSlim with Async/Await

I am not an advanced developer. I'm just trying to get a hold on the task library and just googling. I've never used the class SemaphoreSlim so I would like to know what it does. Here I present code where SemaphoreSlim is used with async & await but which I do not understand. Could someone help me to understand the code below.
1st set of code
await WorkerMainAsync();
async Task WorkerMainAsync()
{
SemaphoreSlim ss = new SemaphoreSlim(10);
while (true)
{
await ss.WaitAsync();
// you should probably store this task somewhere and then await it
var task = DoPollingThenWorkAsync();
}
}
async Task DoPollingThenWorkAsync(SemaphoreSlim semaphore)
{
var msg = Poll();
if (msg != null)
{
await Task.Delay(3000); // process the I/O-bound job
}
// this assumes you don't have to worry about exceptions
// otherwise consider try-finally
semaphore.Release();
}
Firstly, the WorkerMainAsync will be called and a SemaphoreSlim is used. Why is 10 passed to the constructor of SemaphoreSlim?
When does the control come out of the while loop again?
What does ss.WaitAsync(); do?
The DoPollingThenWorkAsync() function is expecting a SemaphoreSlim but is not passed anything when it is called. Is this typo?
Why is await Task.Delay(3000); used?
They could simply use Task.Delay(3000) but why do they use await here instead?
2nd set of code for same purpose
async Task WorkerMainAsync()
{
SemaphoreSlim ss = new SemaphoreSlim(10);
List<Task> trackedTasks = new List<Task>();
while (DoMore())
{
await ss.WaitAsync();
trackedTasks.Add(Task.Run(() =>
{
DoPollingThenWorkAsync();
ss.Release();
}));
}
await Task.WhenAll(trackedTasks);
}
void DoPollingThenWorkAsync()
{
var msg = Poll();
if (msg != null)
{
Thread.Sleep(2000); // process the long running CPU-bound job
}
}
Here is a task & ss.Release added to a list. I really do not understand how tasks can run after adding to a list?
trackedTasks.Add(Task.Run(async () =>
{
await DoPollingThenWorkAsync();
ss.Release();
}));
I am looking forward for a good explanation & help to understand the two sets of code. Thanks
why 10 is passing to SemaphoreSlim constructor.
They are using SemaphoreSlim to limit to 10 tasks at a time. The semaphore is "taken" before each task is started, and each task "releases" it when it finishes. For more about semaphores, see MSDN.
they can use simply Task.Delay(3000) but why they use await here.
Task.Delay creates a task that completes after the specified time interval and returns it. Like most Task-returning methods, Task.Delay returns immediately; it is the returned Task that has the delay. So if the code did not await it, there would be no delay.
just really do not understand after adding task to list how they can run?
In the Task-based Asynchronous Pattern, Task objects are returned "hot". This means they're already running by the time they're returned. The await Task.WhenAll at the end is waiting for them all to complete.

Continuation tasks not executing in correct order

Been trying to execute tasks sequentially but they are executed in a random order instead.
Appending .Unwrap after .ContinueWith doesn't help
Returning a Task of T from these methods instead of Task and assigning their result in the caller doesn't work either
Not sure about signature of my methods, whether they should contain async/await or not.
Sequencing tasks :
Task biographies = LoadArtistBiographies(apiKey);
Task blogs = LoadArtistBlogs(apiKey);
Task familiarity = LoadArtistFamiliarity(apiKey);
Task hottness = LoadArtistHottness(apiKey);
Task images = LoadArtistImages(apiKey);
await biographies.ContinueWith(b => blogs);
await blogs.ContinueWith(f => familiarity);
await familiarity.ContinueWith(h => hottness);
await hottness.ContinueWith(i => images);
await images;
Sample of executed methods :
private async Task LoadArtistBiographies(string apiKey)
{
var parameters = new ArtistBiographiesParameters();
parameters.SetDefaultValues();
parameters.ApiKey = apiKey;
parameters.Id = _artistId;
ArtistBiographies biographies = await Queries.ArtistBiographies(parameters);
ItemsControlBiographies.ItemsSource = biographies.Biographies;
}
The Queries.* methods are also asynchronous :
public static async Task<ArtistBlogs> ArtistBlogs(ArtistBlogsParameters parameters)
What is the correct syntax for chaining tasks that themselves are executing asynchronous tasks ?
If you want to execute the tasks in a specific order, you should await them directly:
await LoadArtistBiographies(apiKey);
await LoadArtistBlogs(apiKey);
await LoadArtistFamiliarity(apiKey);
await LoadArtistHottness(apiKey);
await LoadArtistImages(apiKey);
This will cause the second task (LoadArtistBlogs) to be scheduled after the first task completes.
Right now, the tasks are executing "in random order" because you've assigned them to Task instances, which allows each to be executed simultaneously.
That being said, I would actually recommend changing your methods around to returning the values, instead of assigning them to the datasource within the method:
private async Task<Biographies> LoadArtistBiographiesAsync(string apiKey)
{
var parameters = new ArtistBiographiesParameters();
parameters.SetDefaultValues();
parameters.ApiKey = apiKey;
parameters.Id = _artistId;
var bio = await Queries.ArtistBiographies(parameters);
return bio.Biographies;
}
You could then write these as:
ItemsControlBiographies.ItemsSource = await LoadArtistBiographiesAsync(apiKey);
// Other methods below, with await as this example
This makes the intent as the logic flows through the async methods a bit more clear, in my opinion.
Your example code will start executing all the tasks without waiting for each one to complete. It then waits for them to complete in order.
The key is that an async method starts when you call it. So if you don't want to start it yet, don't call the method yet:
await LoadArtistBiographies(apiKey);
await LoadArtistBlogs(apiKey);
await LoadArtistFamiliarity(apiKey);
await LoadArtistHottness(apiKey);
await LoadArtistImages(apiKey);
await will wait for the given task to complete, it will not start the task. Your Load*-methods all most likely start a task. All five tasks are running in an arbitrary order.
At the point when you get to await, your task may already has finished or not. It does not matter. You call ContinueWith on it, telling your task it should continue with this method once finished. This will return a new Task, on which you finally await.
Actually I've just found a way but without ContinueWith :
ArtistBiographies biographies = await LoadArtistBiographies(apiKey);
ItemsControlBiographies.ItemsSource = biographies.Biographies;
ArtistBlogs blogs = await LoadArtistBlogs(apiKey);
ItemsControlBlogs.ItemsSource = blogs.Blogs;
ArtistFamiliarity familiarity = await LoadArtistFamiliarity(apiKey);
ContentControlFamiliarity.Content = familiarity.artist;
ArtistHotttnesss hottness = await LoadArtistHottness(apiKey);
ContentControlHottness.Content = hottness.Artist;
ArtistImages images = await LoadArtistImages(apiKey);
ItemsControlImages.ItemsSource = images.Images;
Curious if someone could provide the answer using ContinueWith.

Categories