266 lines
8.8 KiB
Rust
266 lines
8.8 KiB
Rust
//! The `broadcast_stage` broadcasts data from a leader node to validators
|
|
//!
|
|
use cluster_info::{ClusterInfo, ClusterInfoError, NodeInfo};
|
|
use counter::Counter;
|
|
use entry::Entry;
|
|
#[cfg(feature = "erasure")]
|
|
use erasure;
|
|
use influx_db_client as influxdb;
|
|
use ledger::Block;
|
|
use log::Level;
|
|
use metrics;
|
|
use packet::SharedBlobs;
|
|
use rayon::prelude::*;
|
|
use result::{Error, Result};
|
|
use service::Service;
|
|
use std::net::UdpSocket;
|
|
use std::sync::atomic::{AtomicBool, AtomicUsize, Ordering};
|
|
use std::sync::mpsc::{Receiver, RecvTimeoutError};
|
|
use std::sync::{Arc, RwLock};
|
|
use std::thread::{self, Builder, JoinHandle};
|
|
use std::time::{Duration, Instant};
|
|
use timing::duration_as_ms;
|
|
use window::{self, SharedWindow, WindowIndex, WindowUtil};
|
|
|
|
#[derive(Debug, PartialEq, Eq, Clone)]
|
|
pub enum BroadcastStageReturnType {
|
|
LeaderRotation,
|
|
ChannelDisconnected,
|
|
}
|
|
|
|
fn broadcast(
|
|
node_info: &NodeInfo,
|
|
broadcast_table: &[NodeInfo],
|
|
window: &SharedWindow,
|
|
receiver: &Receiver<Vec<Entry>>,
|
|
sock: &UdpSocket,
|
|
transmit_index: &mut WindowIndex,
|
|
receive_index: &mut u64,
|
|
) -> Result<()> {
|
|
let id = node_info.id;
|
|
let timer = Duration::new(1, 0);
|
|
let entries = receiver.recv_timeout(timer)?;
|
|
let now = Instant::now();
|
|
let mut num_entries = entries.len();
|
|
let mut ventries = Vec::new();
|
|
ventries.push(entries);
|
|
while let Ok(entries) = receiver.try_recv() {
|
|
num_entries += entries.len();
|
|
ventries.push(entries);
|
|
}
|
|
inc_new_counter_info!("broadcast_stage-entries_received", num_entries);
|
|
|
|
let to_blobs_start = Instant::now();
|
|
let dq: SharedBlobs = ventries
|
|
.into_par_iter()
|
|
.flat_map(|p| p.to_blobs())
|
|
.collect();
|
|
|
|
let to_blobs_elapsed = duration_as_ms(&to_blobs_start.elapsed());
|
|
|
|
// flatten deque to vec
|
|
let blobs_vec: SharedBlobs = dq.into_iter().collect();
|
|
|
|
let blobs_chunking = Instant::now();
|
|
// We could receive more blobs than window slots so
|
|
// break them up into window-sized chunks to process
|
|
let window_size = window.read().unwrap().window_size();
|
|
let blobs_chunked = blobs_vec.chunks(window_size as usize).map(|x| x.to_vec());
|
|
let chunking_elapsed = duration_as_ms(&blobs_chunking.elapsed());
|
|
|
|
let broadcast_start = Instant::now();
|
|
for mut blobs in blobs_chunked {
|
|
let blobs_len = blobs.len();
|
|
trace!("{}: broadcast blobs.len: {}", id, blobs_len);
|
|
|
|
// TODO: move all this into window.rs
|
|
// Index the blobs
|
|
window::index_blobs(node_info, &blobs, receive_index)
|
|
.expect("index blobs for initial window");
|
|
|
|
// keep the cache of blobs that are broadcast
|
|
inc_new_counter_info!("streamer-broadcast-sent", blobs.len());
|
|
{
|
|
let mut win = window.write().unwrap();
|
|
assert!(blobs.len() <= win.len());
|
|
for b in &blobs {
|
|
let ix = b.read().unwrap().get_index().expect("blob index");
|
|
let pos = (ix % window_size) as usize;
|
|
if let Some(x) = win[pos].data.take() {
|
|
trace!(
|
|
"{} popped {} at {}",
|
|
id,
|
|
x.read().unwrap().get_index().unwrap(),
|
|
pos
|
|
);
|
|
}
|
|
if let Some(x) = win[pos].coding.take() {
|
|
trace!(
|
|
"{} popped {} at {}",
|
|
id,
|
|
x.read().unwrap().get_index().unwrap(),
|
|
pos
|
|
);
|
|
}
|
|
|
|
trace!("{} null {}", id, pos);
|
|
}
|
|
for b in &blobs {
|
|
let ix = b.read().unwrap().get_index().expect("blob index");
|
|
let pos = (ix % window_size) as usize;
|
|
trace!("{} caching {} at {}", id, ix, pos);
|
|
assert!(win[pos].data.is_none());
|
|
win[pos].data = Some(b.clone());
|
|
}
|
|
}
|
|
|
|
// Fill in the coding blob data from the window data blobs
|
|
#[cfg(feature = "erasure")]
|
|
{
|
|
erasure::generate_coding(
|
|
&id,
|
|
&mut window.write().unwrap(),
|
|
*receive_index,
|
|
blobs_len,
|
|
&mut transmit_index.coding,
|
|
)?;
|
|
}
|
|
|
|
*receive_index += blobs_len as u64;
|
|
|
|
// Send blobs out from the window
|
|
ClusterInfo::broadcast(
|
|
&node_info,
|
|
&broadcast_table,
|
|
&window,
|
|
&sock,
|
|
transmit_index,
|
|
*receive_index,
|
|
)?;
|
|
}
|
|
let broadcast_elapsed = duration_as_ms(&broadcast_start.elapsed());
|
|
|
|
inc_new_counter_info!(
|
|
"broadcast_stage-time_ms",
|
|
duration_as_ms(&now.elapsed()) as usize
|
|
);
|
|
info!(
|
|
"broadcast: {} entries, blob time {} chunking time {} broadcast time {}",
|
|
num_entries, to_blobs_elapsed, chunking_elapsed, broadcast_elapsed
|
|
);
|
|
|
|
metrics::submit(
|
|
influxdb::Point::new("broadcast-stage")
|
|
.add_field(
|
|
"transmit-index",
|
|
influxdb::Value::Integer(transmit_index.data as i64),
|
|
).to_owned(),
|
|
);
|
|
|
|
Ok(())
|
|
}
|
|
|
|
// Implement a destructor for the BroadcastStage thread to signal it exited
|
|
// even on panics
|
|
struct Finalizer {
|
|
exit_sender: Arc<AtomicBool>,
|
|
}
|
|
|
|
impl Finalizer {
|
|
fn new(exit_sender: Arc<AtomicBool>) -> Self {
|
|
Finalizer { exit_sender }
|
|
}
|
|
}
|
|
// Implement a destructor for Finalizer.
|
|
impl Drop for Finalizer {
|
|
fn drop(&mut self) {
|
|
self.exit_sender.clone().store(true, Ordering::Relaxed);
|
|
}
|
|
}
|
|
|
|
pub struct BroadcastStage {
|
|
thread_hdl: JoinHandle<BroadcastStageReturnType>,
|
|
}
|
|
|
|
impl BroadcastStage {
|
|
fn run(
|
|
sock: &UdpSocket,
|
|
cluster_info: &Arc<RwLock<ClusterInfo>>,
|
|
window: &SharedWindow,
|
|
entry_height: u64,
|
|
receiver: &Receiver<Vec<Entry>>,
|
|
) -> BroadcastStageReturnType {
|
|
let mut transmit_index = WindowIndex {
|
|
data: entry_height,
|
|
coding: entry_height,
|
|
};
|
|
let mut receive_index = entry_height;
|
|
let me = cluster_info.read().unwrap().my_data().clone();
|
|
loop {
|
|
let broadcast_table = cluster_info.read().unwrap().compute_broadcast_table();
|
|
if let Err(e) = broadcast(
|
|
&me,
|
|
&broadcast_table,
|
|
&window,
|
|
&receiver,
|
|
&sock,
|
|
&mut transmit_index,
|
|
&mut receive_index,
|
|
) {
|
|
match e {
|
|
Error::RecvTimeoutError(RecvTimeoutError::Disconnected) => {
|
|
return BroadcastStageReturnType::ChannelDisconnected
|
|
}
|
|
Error::RecvTimeoutError(RecvTimeoutError::Timeout) => (),
|
|
Error::ClusterInfoError(ClusterInfoError::NoPeers) => (), // TODO: Why are the unit-tests throwing hundreds of these?
|
|
_ => {
|
|
inc_new_counter_info!("streamer-broadcaster-error", 1, 1);
|
|
error!("broadcaster error: {:?}", e);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
/// Service to broadcast messages from the leader to layer 1 nodes.
|
|
/// See `cluster_info` for network layer definitions.
|
|
/// # Arguments
|
|
/// * `sock` - Socket to send from.
|
|
/// * `exit` - Boolean to signal system exit.
|
|
/// * `cluster_info` - ClusterInfo structure
|
|
/// * `window` - Cache of blobs that we have broadcast
|
|
/// * `receiver` - Receive channel for blobs to be retransmitted to all the layer 1 nodes.
|
|
/// * `exit_sender` - Set to true when this stage exits, allows rest of Tpu to exit cleanly. Otherwise,
|
|
/// when a Tpu stage closes, it only closes the stages that come after it. The stages
|
|
/// that come before could be blocked on a receive, and never notice that they need to
|
|
/// exit. Now, if any stage of the Tpu closes, it will lead to closing the WriteStage (b/c
|
|
/// WriteStage is the last stage in the pipeline), which will then close Broadcast stage,
|
|
/// which will then close FetchStage in the Tpu, and then the rest of the Tpu,
|
|
/// completing the cycle.
|
|
pub fn new(
|
|
sock: UdpSocket,
|
|
cluster_info: Arc<RwLock<ClusterInfo>>,
|
|
window: SharedWindow,
|
|
entry_height: u64,
|
|
receiver: Receiver<Vec<Entry>>,
|
|
exit_sender: Arc<AtomicBool>,
|
|
) -> Self {
|
|
let thread_hdl = Builder::new()
|
|
.name("solana-broadcaster".to_string())
|
|
.spawn(move || {
|
|
let _exit = Finalizer::new(exit_sender);
|
|
Self::run(&sock, &cluster_info, &window, entry_height, &receiver)
|
|
}).unwrap();
|
|
|
|
BroadcastStage { thread_hdl }
|
|
}
|
|
}
|
|
|
|
impl Service for BroadcastStage {
|
|
type JoinReturnType = BroadcastStageReturnType;
|
|
|
|
fn join(self) -> thread::Result<BroadcastStageReturnType> {
|
|
self.thread_hdl.join()
|
|
}
|
|
}
|