Consider the following two ways of constructing an Async computation that calculates 1 + 2:
let c1 =
async {
let a = 1
let b = 2
return a + b }
let c2 =
async {
let a = 1
let! b = async { return 2 }
return a + b }
What is the practical difference between them? It seems to me that they do the same thing. Why, for example, would you ever need to use let! result = streamReader.ReadToEndAsync () rather than let result = streamReader.ReadToEnd ()? Isn't it the case that both lines are "blocking" when the computation is run?
Your silly example with let! b = async { return 2 } is indeed not bringing anything new. It's indeed completely equivalent to let b = 2.
It is a different story, however, with ReadToEnd vs. ReadToEndAsync. While they can both be described as "blocking", they are blocking is quite different ways.
ReadToEnd is synchronous. When it's called on a thread, that thread stops and waits for it to complete. The thread is blocked. It's doing nothing, but it also can't be used to execute anything else.
ReadToEndAsync uses asynchronous I/O (also called "overlapped I/O" on Windows). This basically means that the thread stops at this point and calls the OS, saying "hey, please read this file for me and wake me up when you're done". The way it's implemented at a lower level differs depending on the OS, but generally you get some sort of callback when the call completes.
This is kind of important in high-availability high-concurrency systems, such as HTTP servers. But if you're just running a script locally on your computer under human supervision, then just use whatever is more convenient.
Interestingly, the sync version ReadToEnd does actually use async I/O under the hood as well, it's just wrapped in a sync-blocking wrapper to make it more convenient to use in simple cases.
Related
Using the futures crate.
I have a vec of futures which return a bool and I want to wait specifically for the future that returns true.
consider the following pool of futures.
async fn async_function(guess: u8) -> bool {
let random_wait = rand::thread_rng().gen_range(0..2);
std::thread::sleep(std::time::Duration::from_secs(random_wait));
println!("Running guess {guess}");
guess == 231
}
fn main() {
let mut pool: Vec<impl Future<Output = bool>> = vec![];
for guess in 0..=255 {
pool.push(async_function(guess));
}
}
How do I wait for the futures in the vec?
Is it possible to wait until only one future returns true?
Can I identify the value of guess for the future that returns true?
I'm new to async rust, so I've been looking at the async-book.
From there, these are the options I've considered:
join! waits until all threads are done, so that doesn't work for me since I want to drop the remaining futures.
select! doesn't seem to be an option, because I need to specify the specific future in the select block and I'm not about to make 255 line select.
try_join! is tempting me to break semantics and have my async_function return Err(guess)so that it causes the try_join to exit and return the value I want.
I tried using async_fn(guess).boxed.into_stream() and then using select_all from futures::stream but it doesn't seem to run concurrently. I see my async functions running in order.
Ok, my thinking of futures was wrong. I knew that they weren't executed immediately, but I wasn't using the executors correctly from the futures crate.
Here's what I've got that seems to work.
let thread_pool = ThreadPool::new().unwrap();
let mut pool = vec![];
for guess in 0..=255 {
let thread = thread_pool.spawn_with_handle(async_fn(guess)).expect("Failed to spawn thread");
pool.push(thread.into_stream());
}
let stream = block_on_stream(futures::stream::select_all(pool));
for value in stream {
println!("Got value {value}");
}
the thread pool executor is what creates the separate threads needed to run. Without this my application was single threaded so no matter what I tried, it would only run functions one at a time, not concurrently.
This way I spawn each request into the thread pool. The thread pool appears to spawn 4 background threads. By pushing them all into a stream, using select_all, and iterating over the stream, my main thread blocks until a new value is available.
There's always 4 workers and the thread pool is scheduling them in the order they were requested like a queue. This is perfect.
I just started learning asynchronous Rust, so this is propably not a difficult question to answer, however, I am scratching my head here.
I am not trying to run tasks in parallel yet, only trying to get them to run concurrently.
According to the guide at https://rust-lang.github.io/async-book/,
The futures::join macro makes it possible to wait for multiple different futures to complete while executing them all concurrently.
So when I create 2 Futures, I should be able to "await" both of them at once. It also states that
Whereas calling a blocking function in a synchronous method would block the whole thread, blocked Futures will yield control of the thread, allowing other Futures to run.
From what I understand here, if I await multiple Futures with join!, should the first one be blocked, the second one will start running.
So I made a very simple example where I created 2 async fns and tried to join! both, making sure the first one gets blocked. I used a mpsc::channel for the blocking, since the docs stated that thread::sleep() should not be used in async fns and that recv()
will always block the current thread if there is no data available
However, the behavior is not what I expected, as calling the blocking function will not yield control of the thread, allowing the other Future to run, like I would expect from the second quote I provided. Instead, it will just wait untill it is no longer blocked, finish the first Future and only then start the second. Pretty much as if they were synchronous and I would have just called one after the other.
My complete example code:
use std::{thread::{self}, sync::{mpsc::{self, Sender, Receiver}}, time::Duration};
use futures::{executor}; //added futures = "0.3" in cargo.toml dependencies
fn main(){
let fut = main_async();
executor::block_on(fut);
}
async fn main_async(){
let (sender, receiver) = mpsc::channel();
let thread_handle = std::thread::spawn(move || { //this thread is just here so the f1 function gets blocked by something and can later resume
wait_send_function(sender);
});
let f1 = f1(receiver);
let f2 = f2();
futures::join!(f1, f2);
thread_handle.join().unwrap();
}
fn wait_send_function(sender: Sender<i32>){
thread::sleep(Duration::from_millis(5000));
sender.send(1234).unwrap();
}
async fn f1(receiver: Receiver<i32>){
println!("starting f1");
let new_nmbr = receiver.recv().unwrap(); //I would expect f2 to start now, since this is blocking
println!("Received nmbr is: {}", new_nmbr);
}
async fn f2(){
println!("starting f2");
}
And the output is simply:
starting f1
Received nmbr is: 1234
starting f2
My question is what am I missing here, why does f2 only start after f1 is completed and what would I need to do to get the behavior I want (completing f2 first if f1 is blocked and then waiting for f1)?
Maybe the book is a little misleading, but when it refers to "a blocked future", it does not mean in the sense of blocking synchronous code (if that was the case, there would be no problem to use std::thread::sleep()), but rather, it means that the future is waiting to be polled by the executor.
Thus, std::mpsc that blocks the thread will not have the desired effect (definitely not on a single-threaded executor like future's, but it's a bad idea on multi-threaded executors too). Use futures::channel::mpsc and everything will work.
Boost's asio library allows the serialisation of asynchronous code in the following way. Handlers to asynchronous functions such as those which read from a stream, may be associated to a strand. A strand is associated with an "IO context". An IO context owns a thread pool. However many threads in the pool, it is guaranteed that no two handlers associated with the same strand are run concurrently. This makes it possible, for instance, to implement a state machine as if it were single-threaded, where all handlers for that machine serialise over a private strand.
I have been trying to figure out how this might be done with F#'s Async. I could not find any way to make sure that chosen sets of Async processes never run concurrently. Can anyone suggest how to do this?
It would be useful to know what is the use case that you are trying to implement. I don't think F# async has anything that would directly map to strands and you would likely use different techniques for implementing different things that might all be implemented using strands.
For example, if you are concerend with reading data from a stream, F# async block lets you write code that is asynchronous but sequential. The following runs a single logical process (which might be moved between threads of a thread pool when you wait using let!):
let readTest () = async {
let fs = File.OpenRead(#"C:\Temp\test.fs")
let buffer = Array.zeroCreate 10
let mutable read = 1
while read <> 0 do
let! r = fs.AsyncRead(buffer, 0, 10)
printfn "Read: %A" buffer.[0 .. r-1]
read <- r }
readTest() |> Async.Start
If you wanted to deal with events that occur without any control (i.e. push based rather than pull based), for example, when you cannot ask the system to read next buffer of data, you could serialize the events using a MailboxProcessor. The following sends two messages to the agent almost at the same time, but they are processed sequentially, with 1 second delay:
let agent = MailboxProcessor.Start(fun inbox -> async {
while true do
let! msg = inbox.Receive()
printfn "Got: %s" msg
do! Async.Sleep(1000)
})
agent.Post("hello")
agent.Post("world")
I have a C# API like this:
Task<T> Foo(serverUri)
Let's say I have 4 possible serverUris. I want to implement a function that will return DiscUnionBar type:
type DiscUnionBar of T =
Safe of T | Weak of T | ConnectionError
The implementation will have the following requirements:
Do 3 (max) concurrent calls to Foo() with 3 different serverUris.
Pick the 2 fastest successful responses. If they give same result T1 and T2 (being T1==T2), stop doing concurrent requests and ignore/cancel requests that are in progress and return Safe of T. If T1!=T2, keep doing more requests (or looking at responses) until two equal responses are found.
If any of the requests fails (throws ServerException), try with a serverUri that has not been requested before.
If all requests to all 4 servers fail, return ConnectionError.
If only 1 request succeeds, return Weak of T.
Is this easy to do given that I cannot use F#'s Async and have to stick with C#'s Task usage? I'm a bit lost on this one.
Unless there is a reason you cannot use Async anywhere in your code, and your only limitation is that Foo has to return a Task, you should have no problem converting the Task resulting from calling Foo to an Async with Async.AwaitTask.
This way you can build the logic using F#'s async computation expressions as if Foo returned an Async
let simpleComputation serverUri = async {
let! fooResult = Foo(serverUri) |> Async.AwaitTask
(* here you can work with the T returned by Foo's task *)
}
I also have good experience with FSharp.Control.FusionTasks library, which lets you use Task directly in async computation expressions, without having to call AwaitTask explicitly, and helps in Async/Task interop in general. Although some may not like that it tries to hide the Tasks.
To execute operations on a background thread and avoid blocking the UI in a WPF application, I often find myself writing this pattern:
async {
// some code on the UI thread
let uiThread = SynchronizationContext.Current
do! Async.SwitchToThreadPool()
let! result = // some Async<'t>
do! Async.SwitchToContext uiThread
// do things with the result if it wasn't () all along
}
Am I doing this right at all? Is this idiomatic? Should it be done differently?
If this is correct, of course I would prefer not to have to do it like that all the time - is there a built-in shorter way to achieve the same thing? None of the existing Async functions appears to do something like that.
If not, does it make sense to just turn the above code into a function?
let onThreadPool operation =
async {
let context = SynchronizationContext.Current
do! Async.SwitchToThreadPool()
let! result = operation
do! Async.SwitchToContext context
return result
}
That adds another level of async { } nesting - can this cause issues at "some" point?
What you're doing here definitely makes sense. One useful operation here is Async.StartImmediate, which starts the async workflow on the current thread. If you call this from the UI thread, this guarantees that the workflow will also start on the UI thread and so you can capture the synchronization context inside the workflow.
The other trick is that many built-in asynchronous F# operations automatically jump back to the original synchronization context (those that are created using Async.FromContinuations, including e.g. AsyncDownloadString), so when you're calling one of those, you do not even need to explicitly jump back to the original synchronization context.
But for other asynchronous operations (and for non-async operations that you want to run in the background), your onThreadPool function looks like a great way of doing this.
#random-dev is right capturing the context must happen outside the workflow
let onThreadPool operation =
let context = SynchronizationContext.Current
async {
do! Async.SwitchToThreadPool()
let! result = operation
do! Async.SwitchToContext context
return result
}