Improve the chat example, making it more robust (#199)

This handles cases where clients send large amounts of data while on
localhost.

Closes #192
This commit is contained in:
Carl Lerche
2018-03-08 13:38:52 -08:00
committed by GitHub
parent 142bd3b2a9
commit fed4d72eff
+25 -1
View File
@@ -172,8 +172,20 @@ impl Future for Peer {
type Error = io::Error;
fn poll(&mut self) -> Poll<(), io::Error> {
// Tokio (and futures) use cooperative scheduling without any
// preemption. If a task never yields execution back to the executor,
// then other tasks may be starved.
//
// To deal with this, robust applications should not have any unbounded
// loops. In this example, we will read at most `LINES_PER_TICK` lines
// from the client on each tick.
//
// If the limit is hit, the current task is notified, informing the
// executor to schedule the task again asap.
const LINES_PER_TICK: usize = 10;
// Receive all messages from peers.
loop {
for i in 0..LINES_PER_TICK {
// Polling an `UnboundedReceiver` cannot fail, so `unwrap` here is
// safe.
match self.rx.poll().unwrap() {
@@ -181,6 +193,14 @@ impl Future for Peer {
// Buffer the line. Once all lines are buffered, they will
// be flushed to the socket (right below).
self.lines.buffer(&v);
// If this is the last iteration, the loop will break even
// though there could still be lines to read. Because we did
// not reach `Async::NotReady`, we have to notify ourselves
// in order to tell the executor to schedule the task again.
if i+1 == LINES_PER_TICK {
task::current().notify();
}
}
_ => break,
}
@@ -256,6 +276,10 @@ impl Lines {
/// This writes the line to an internal buffer. Calls to `poll_flush` will
/// attempt to flush this buffer to the socket.
fn buffer(&mut self, line: &[u8]) {
// Ensure the buffer has capacity. Ideally this would not be unbounded,
// but to keep the example simple, we will not limit this.
self.wr.reserve(line.len());
// Push the line onto the end of the write buffer.
//
// The `put` function is from the `BufMut` trait.