Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

test: restore unstable network tests #5197

Draft
wants to merge 4 commits into
base: main
Choose a base branch
from
Draft
Show file tree
Hide file tree
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Next Next commit
feat: implement p2p TCP relay
Signed-off-by: 0x009922 <[email protected]>
  • Loading branch information
0x009922 committed Nov 5, 2024
commit b7932c347226d4f2cc08337a5a5ac10f89e83e52
7 changes: 7 additions & 0 deletions Cargo.lock

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

1 change: 1 addition & 0 deletions crates/iroha/Cargo.toml
Original file line number Diff line number Diff line change
Expand Up @@ -95,3 +95,4 @@ hex = { workspace = true }
assertables = { workspace = true }
trybuild = { workspace = true }
assert_matches = "1.5.0"
ascii_table = "4.0.4"
363 changes: 363 additions & 0 deletions crates/iroha/tests/faulty_peers.rs
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What do you think about renaming "faulty peers" to "fault tolerance", which sounds to me like a clearer expression of what this test should guarantee

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sounds good to me.

Original file line number Diff line number Diff line change
@@ -0,0 +1,363 @@
use std::time::Duration;

use eyre::Result;
use futures_util::{stream::FuturesUnordered, StreamExt};
use iroha_config_base::toml::WriteExt;
use iroha_test_network::{
genesis_factory, once_blocks_sync, Network, NetworkBuilder, PeerLifecycleEvent,
};
use relay::P2pRelay;
use tokio::{self, time::timeout};

mod relay {
use std::{
collections::HashMap,
iter::once,
sync::{
atomic::{AtomicBool, Ordering},
Arc,
},
};

use futures_util::{stream::FuturesUnordered, StreamExt};
use iroha_data_model::peer::PeerId;
use iroha_primitives::{
addr::{socket_addr, SocketAddr},
unique_vec::UniqueVec,
};
use iroha_test_network::fslock_ports::AllocatedPort;
use tokio::{
io::{AsyncRead, AsyncReadExt, AsyncWrite, AsyncWriteExt},
net::{TcpListener, TcpStream},
select,
sync::Notify,
task::JoinSet,
};

#[derive(Debug)]
pub struct P2pRelay {
peers: HashMap<PeerId, RelayPeer>,
tasks: JoinSet<()>,
}

#[derive(Debug)]
struct RelayPeer {
real_addr: SocketAddr,
mock_outgoing: HashMap<PeerId, (SocketAddr, AllocatedPort)>,
suspend: Suspend,
}

impl P2pRelay {
pub fn new(real_topology: &UniqueVec<PeerId>) -> Self {
let peers: HashMap<_, _> = real_topology
.iter()
.map(|peer_id| {
let real_addr = peer_id.address().clone();
let mock_outgoing = real_topology
.iter()
.filter(|x| *x != peer_id)
.map(|other_id| {
let mock_port = AllocatedPort::new();
let mock_addr = socket_addr!(127.0.0.1:*mock_port);
(other_id.clone(), (mock_addr, mock_port))
})
.collect();
let peer = RelayPeer {
real_addr,
mock_outgoing,
suspend: Suspend::new(),
};
(peer_id.clone(), peer)
})
.collect();

let mut table = ascii_table::AsciiTable::default();
table.set_max_width(30 * (1 + real_topology.len()));
table.column(0).set_header("From");
for (i, id) in real_topology.iter().enumerate() {
table
.column(i + 1)
.set_header(format!("To {}", id.address()));
}
table.print(real_topology.iter().map(|id| {
once(format!("{}", id.address()))
.chain(real_topology.iter().map(|peer_id| {
if *peer_id == *id {
"".to_string()
} else {
let (mock_addr, _) =
peers.get(id).unwrap().mock_outgoing.get(peer_id).unwrap();
format!("{mock_addr}")
}
}))
.collect::<Vec<_>>()
}));

Self {
peers,
tasks: <_>::default(),
}
}

pub fn topology_for(&self, peer: &PeerId) -> UniqueVec<PeerId> {
self.peers
.get(peer)
.expect("existing peer must be supplied")
.mock_outgoing
.iter()
.map(|(other, (addr, _port))| PeerId::new(addr.clone(), other.public_key().clone()))
.collect()
}

pub fn start(&mut self) {
for (_peer_id, peer) in self.peers.iter() {
for (other_id, (other_mock_addr, _)) in peer.mock_outgoing.iter() {
let other_peer = self.peers.get(other_id).expect("must be present");
let suspend =
SuspendIfAny(vec![peer.suspend.clone(), other_peer.suspend.clone()]);

P2pRelay::run_proxy(
&mut self.tasks,
other_mock_addr.clone(),
other_peer.real_addr.clone(),
suspend,
);
}
}
}

fn run_proxy(
tasks: &mut JoinSet<()>,
from: SocketAddr,
to: SocketAddr,
suspend: SuspendIfAny,
) {
eprintln!("proxy: {from} → {to}");
let mut proxy = Proxy::new(from, to, suspend);

tasks.spawn(async move {
if let Err(err) = proxy.run().await {
eprintln!("proxy at {} exited with an error: {err}", proxy.from);
} else {
eprintln!("proxy exited normally");
}
});
}

pub fn suspend(&self, peer: &PeerId) -> Suspend {
self.peers
.get(peer)
.expect("must be present")
.suspend
.clone()
}
}

#[derive(Clone, Debug, Default)]
pub struct Suspend {
active: Arc<AtomicBool>,
notify: Arc<Notify>,
}

impl Suspend {
fn new() -> Self {
Self::default()
}

pub fn activate(&self) {
self.active.store(true, Ordering::Release);
}

pub fn deactivate(&self) {
self.active.store(false, Ordering::Release);
self.notify.notify_waiters();
}
}

#[derive(Clone, Debug)]
struct SuspendIfAny(Vec<Suspend>);

impl SuspendIfAny {
async fn is_not_active(&self) {
loop {
let waited_for = self
.0
.iter()
.filter_map(|x| {
x.active
.load(Ordering::Acquire)
.then_some(x.notify.notified())
})
.collect::<FuturesUnordered<_>>()
.collect::<Vec<_>>()
.await
.len();
if waited_for == 0 {
break;
}
}
}
}

struct Proxy {
from: SocketAddr,
to: SocketAddr,
suspend: SuspendIfAny,
}

impl Proxy {
fn new(from: SocketAddr, to: SocketAddr, suspend: SuspendIfAny) -> Self {
Self { from, to, suspend }
}

async fn run(&mut self) -> eyre::Result<()> {
let listener = TcpListener::bind(self.from.to_string()).await?;
loop {
let (client, _) = listener.accept().await?;
let server = TcpStream::connect(self.to.to_string()).await?;

let (mut eread, mut ewrite) = client.into_split();
let (mut oread, mut owrite) = server.into_split();

let suspend = self.suspend.clone();
let e2o =
tokio::spawn(
async move { Proxy::copy(&suspend, &mut eread, &mut owrite).await },
);
let suspend = self.suspend.clone();
let o2e =
tokio::spawn(
async move { Proxy::copy(&suspend, &mut oread, &mut ewrite).await },
);

select! {
_ = e2o => {
// eprintln!("{} → {}: client-to-server closed ×", self.from, self.to);
},
_ = o2e => {
// eprintln!("{} → {}: server-to-client closed ×", self.from, self.to);
},
}
}
}

async fn copy<R, W>(
suspend: &SuspendIfAny,
mut reader: R,
mut writer: W,
) -> std::io::Result<()>
where
R: AsyncRead + Unpin,
W: AsyncWrite + Unpin,
{
// NOTE: stack overflow happens without the box
let mut buf = Box::new([0u8; 2usize.pow(20)]);

loop {
suspend.is_not_active().await;

let n = reader.read(&mut *buf).await?;
if n == 0 {
break;
}

writer.write_all(&buf[..n]).await?;
}

Ok(())
}
}
}

async fn start_network_with_relay(network: &Network) -> Result<P2pRelay> {
let relay = P2pRelay::new(&network.peers().iter().map(|peer| peer.id()).collect());

timeout(
network.peer_startup_timeout(),
network
.peers()
.iter()
.enumerate()
.map(|(i, peer)| {
let topology = relay.topology_for(&peer.id());
let config = network
.config()
.write(["sumeragi", "trusted_peers"], &topology);
let genesis = genesis_factory(network.genesis_isi().clone(), topology);
async move {
// FIXME: parallel
peer.start(config, (i == 0).then_some(&genesis)).await;
peer.once(|e| matches!(e, PeerLifecycleEvent::ServerStarted))
.await;
}
})
.collect::<FuturesUnordered<_>>()
.collect::<Vec<_>>(),
)
.await?;

Ok(relay)
}

#[tokio::test]
async fn network_starts_with_relay() -> Result<()> {
let network = NetworkBuilder::new().with_peers(4).build();
let mut relay = start_network_with_relay(&network).await?;

relay.start();
network.ensure_blocks(1).await?;

Ok(())
}

#[tokio::test]
async fn network_doesnt_start_without_relay() -> Result<()> {
let network = NetworkBuilder::new().with_peers(4).build();
let _relay = start_network_with_relay(&network).await?;

if let Ok(_) = timeout(
Duration::from_secs(3),
once_blocks_sync(network.peers().iter(), 1),
)
.await
{
panic!("network must not start!")
};

Ok(())
}

#[tokio::test]
async fn suspending_works() -> Result<()> {
const SYNC: Duration = Duration::from_secs(3);
const N_PEERS: usize = 4;
const { assert!(N_PEERS > 0) };

let network = NetworkBuilder::new().with_peers(N_PEERS).build();
let mut relay = start_network_with_relay(&network).await?;
// we will plug/unplug the last peer who doesn't have the genesis
let last_peer = network
.peers()
.last()
.expect("there are more than 0 of them");
let suspend = relay.suspend(&last_peer.id());

suspend.activate();
relay.start();

// all peers except the last one should get the genesis
timeout(
SYNC,
once_blocks_sync(network.peers().iter().take(N_PEERS - 1), 1),
)
.await?;
let Err(_) = timeout(SYNC, last_peer.once_block(1)).await else {
panic!("should not get block within timeout!")
};

// unsuspend, the last peer should get the block too
suspend.deactivate();
timeout(SYNC, last_peer.once_block(1)).await?;

Ok(())
}
Loading