Skip to content
Branch: master
Find file Copy path
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
100 lines (81 sloc) 3.85 KB
use tokio::runtime::Builder;
use crossbeam::crossbeam_channel::{bounded, unbounded, Receiver, Sender};
use std::thread;
fn main() -> Result<(), Box<dyn std::error::Error>> {
// Runtime for IO related tasks, give few cores for it
let rt_fetching = Builder::new()
// Runtime for CPU intensive tasks, give more cores for it
let rt_analyzing = Builder::new()
// Channels to pass on fetching (IO task) results
let (chan_pages_s, chan_pages_r) = unbounded();
// Channels to pass on analyzing (CPU task) results
let (chan_data_s, chan_data_r) = bounded(16);
rt_fetching.block_on(async {
// simulating multiple IO fetch tasks, each one assign a worker, lots of workers go fetching
for i in 1..=100 {
tokio::spawn(get_content(i.to_string(), chan_pages_s.clone()));
rt_analyzing.block_on(async {
// simulating 10 workers doing CPU intensive tasks
for _ in 0..10 {
tokio::spawn(analyze_content(chan_pages_r.clone(), chan_data_s.clone()));
println!("Hello, world!");
drop(chan_data_s); // close sender from current thread, so to allow collecting on receiver to continue (when other senders in other threads are dropped)
let result_list: Vec<usize> = chan_data_r.iter().collect();
println!("number of result_list: {}", result_list.len());
let sum: usize = result_list.iter().sum::<usize>();
println!("sum of result_list: {}", sum);
async fn get_content(n: String, sender: Sender<String>) {
// TODO: later on use some .await call to get real pages and then send content to sender
// We can also simulate some CPU delay here
// Resulting in each worker is delayed by 1s, 2 thread, 100 workers
// with 100 data, should be finished within 1 seconds.
let simulate_delay = tokio::clock::now() + std::time::Duration::from_secs(1);
// If you try below with blocking the actual thread, the worker thread will be blocked
// resulting only 2 tasks can be done for each second, as we have only 2 thread.
// thread::sleep(std::time::Duration::from_secs(1));
// Which means, in order for this async IO task works well, all internal
// call here in this function should be async (with .await) and not making it blocking!
"[{}]: Async fetch and send page content {}...",
sender.send(n).expect("get_content send ok");
async fn analyze_content(receiver: Receiver<String>, sender: Sender<usize>) {
while let Ok(page) = receiver.recv() {
// TODO: later on switch this to real CPU intensive tasks
// for now, only switching a string to usize as some analyze function
let data: usize = page.parse().unwrap_or(0);
// We can also simulate some CPU delay here
// Resulting in each thread is delayed by 1s, 10 workers, 10 block thread,
// with 100 data, should be finished within 10 seconds.
let simulate_delay = tokio::clock::now() + std::time::Duration::from_secs(1);
// As we have 10 workers and 10 thread, this would be the same effect above
// thread::sleep(std::time::Duration::from_secs(1));
// Thus, by design this method call can have CPU blocking call.
"[{}]: Async analysis and send data content {}...",
sender.send(data).expect("analyze_content send ok");
You can’t perform that action at this time.