Merge branch 'dht-performance' into 'main'

RPC improvements

See merge request veilid/veilid!361
This commit is contained in:
Christien Rioux 2025-03-05 11:14:31 -05:00
commit d0e751ea25
24 changed files with 417 additions and 214 deletions

View file

@ -289,3 +289,91 @@ macro_rules! veilid_log {
$($k).+ = $($fields)* $($k).+ = $($fields)*
)}; )};
} }
#[macro_export]
macro_rules! network_result_value_or_log {
($self:ident $r:expr => $f:expr) => {
network_result_value_or_log!($self target: self::__VEILID_LOG_FACILITY, $r => [ "" ] $f )
};
($self:ident $r:expr => [ $d:expr ] $f:expr) => {
network_result_value_or_log!($self target: self::__VEILID_LOG_FACILITY, $r => [ $d ] $f )
};
($self:ident target: $target:expr, $r:expr => $f:expr) => {
network_result_value_or_log!($self target: $target, $r => [ "" ] $f )
};
($self:ident target: $target:expr, $r:expr => [ $d:expr ] $f:expr) => { {
let __extra_message = if debug_target_enabled!("network_result") {
$d.to_string()
} else {
"".to_string()
};
match $r {
NetworkResult::Timeout => {
veilid_log!($self debug target: $target,
"{} at {}@{}:{} in {}{}",
"Timeout",
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::ServiceUnavailable(ref s) => {
veilid_log!($self debug target: $target,
"{}({}) at {}@{}:{} in {}{}",
"ServiceUnavailable",
s,
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::NoConnection(ref e) => {
veilid_log!($self debug target: $target,
"{}({}) at {}@{}:{} in {}{}",
"No connection",
e.to_string(),
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::AlreadyExists(ref e) => {
veilid_log!($self debug target: $target,
"{}({}) at {}@{}:{} in {}{}",
"Already exists",
e.to_string(),
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::InvalidMessage(ref s) => {
veilid_log!($self debug target: $target,
"{}({}) at {}@{}:{} in {}{}",
"Invalid message",
s,
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::Value(v) => v,
}
} };
}

View file

@ -8,7 +8,7 @@ impl_veilid_log_facility!("net");
const PROTECTED_CONNECTION_DROP_SPAN: TimestampDuration = TimestampDuration::new_secs(10); const PROTECTED_CONNECTION_DROP_SPAN: TimestampDuration = TimestampDuration::new_secs(10);
const PROTECTED_CONNECTION_DROP_COUNT: usize = 3; const PROTECTED_CONNECTION_DROP_COUNT: usize = 3;
const NEW_CONNECTION_RETRY_COUNT: usize = 1; const NEW_CONNECTION_RETRY_COUNT: usize = 0;
const NEW_CONNECTION_RETRY_DELAY_MS: u32 = 500; const NEW_CONNECTION_RETRY_DELAY_MS: u32 = 500;
/////////////////////////////////////////////////////////// ///////////////////////////////////////////////////////////
@ -415,7 +415,19 @@ impl ConnectionManager {
let best_port = preferred_local_address.map(|pla| pla.port()); let best_port = preferred_local_address.map(|pla| pla.port());
// Async lock on the remote address for atomicity per remote // Async lock on the remote address for atomicity per remote
let _lock_guard = self.arc.address_lock_table.lock_tag(remote_addr).await; // Use the initial connection timeout here because multiple calls to get_or_create_connection
// can be performed simultaneously and we want to wait for the first one to succeed or not
let Ok(_lock_guard) = timeout(
self.arc.connection_initial_timeout_ms,
self.arc.address_lock_table.lock_tag(remote_addr),
)
.await
else {
veilid_log!(self debug "== get_or_create_connection: connection busy, not connecting to dial_info={:?}", dial_info);
return Ok(NetworkResult::no_connection_other(
"connection endpoint busy",
));
};
veilid_log!(self trace "== get_or_create_connection dial_info={:?}", dial_info); veilid_log!(self trace "== get_or_create_connection dial_info={:?}", dial_info);
@ -449,7 +461,8 @@ impl ConnectionManager {
let mut retry_count = NEW_CONNECTION_RETRY_COUNT; let mut retry_count = NEW_CONNECTION_RETRY_COUNT;
let network_manager = self.network_manager(); let network_manager = self.network_manager();
let prot_conn = network_result_try!(loop { let nres = loop {
veilid_log!(self trace "== get_or_create_connection connect({}) {:?} -> {}", retry_count, preferred_local_address, dial_info);
let result_net_res = ProtocolNetworkConnection::connect( let result_net_res = ProtocolNetworkConnection::connect(
self.registry(), self.registry(),
preferred_local_address, preferred_local_address,
@ -474,12 +487,16 @@ impl ConnectionManager {
} }
} }
}; };
veilid_log!(self debug "get_or_create_connection retries left: {}", retry_count);
retry_count -= 1; retry_count -= 1;
// Release the preferred local address if things can't connect due to a low-level collision we dont have a record of // // XXX: This should not be necessary
preferred_local_address = None; // // Release the preferred local address if things can't connect due to a low-level collision we dont have a record of
// preferred_local_address = None;
sleep(NEW_CONNECTION_RETRY_DELAY_MS).await; sleep(NEW_CONNECTION_RETRY_DELAY_MS).await;
};
let prot_conn = network_result_value_or_log!(self target:"network_result", nres => [ format!("== get_or_create_connection failed {:?} -> {}", preferred_local_address, dial_info) ] {
network_result_raise!(nres);
}); });
// Add to the connection table // Add to the connection table
@ -598,7 +615,7 @@ impl ConnectionManager {
// Callback from network connection receive loop when it exits // Callback from network connection receive loop when it exits
// cleans up the entry in the connection table // cleans up the entry in the connection table
pub(super) async fn report_connection_finished(&self, connection_id: NetworkConnectionId) { pub(super) fn report_connection_finished(&self, connection_id: NetworkConnectionId) {
// Get channel sender // Get channel sender
let sender = { let sender = {
let mut inner = self.arc.inner.lock(); let mut inner = self.arc.inner.lock();
@ -668,7 +685,7 @@ impl ConnectionManager {
} }
} }
} }
let _ = sender.send_async(ConnectionManagerEvent::Dead(conn)).await; let _ = sender.send(ConnectionManagerEvent::Dead(conn));
} }
} }

View file

@ -662,11 +662,9 @@ impl Network {
}; };
} else { } else {
// Handle connection-oriented protocols // Handle connection-oriented protocols
let connmgr = self.network_manager().connection_manager();
let conn = network_result_try!( let conn = network_result_try!(
self.network_manager() connmgr.get_or_create_connection(dial_info.clone()).await?
.connection_manager()
.get_or_create_connection(dial_info.clone())
.await?
); );
if let ConnectionHandleSendResult::NotSent(_) = conn.send_async(data).await { if let ConnectionHandleSendResult::NotSent(_) = conn.send_async(data).await {

View file

@ -6,7 +6,7 @@ use async_tungstenite::tungstenite::handshake::server::{
Callback, ErrorResponse, Request, Response, Callback, ErrorResponse, Request, Response,
}; };
use async_tungstenite::tungstenite::http::StatusCode; use async_tungstenite::tungstenite::http::StatusCode;
use async_tungstenite::tungstenite::protocol::{frame::coding::CloseCode, CloseFrame, Message}; use async_tungstenite::tungstenite::protocol::Message;
use async_tungstenite::tungstenite::Error; use async_tungstenite::tungstenite::Error;
use async_tungstenite::{accept_hdr_async, client_async, WebSocketStream}; use async_tungstenite::{accept_hdr_async, client_async, WebSocketStream};
use futures_util::{AsyncRead, AsyncWrite, SinkExt}; use futures_util::{AsyncRead, AsyncWrite, SinkExt};
@ -98,45 +98,27 @@ where
#[instrument(level = "trace", target = "protocol", err, skip_all)] #[instrument(level = "trace", target = "protocol", err, skip_all)]
pub async fn close(&self) -> io::Result<NetworkResult<()>> { pub async fn close(&self) -> io::Result<NetworkResult<()>> {
let timeout_ms = self
.registry
.config()
.with(|c| c.network.connection_initial_timeout_ms);
// Make an attempt to close the stream normally // Make an attempt to close the stream normally
let mut stream = self.stream.clone(); let mut stream = self.stream.clone();
let out = match stream
.send(Message::Close(Some(CloseFrame {
code: CloseCode::Normal,
reason: "".into(),
})))
.await
{
Ok(v) => NetworkResult::value(v),
Err(e) => err_to_network_result(e),
};
// This close does not do a TCP shutdown so it is safe and will not cause TIME_WAIT // This close does not do a TCP shutdown so it is safe and will not cause TIME_WAIT
let _ = stream.close().await; match timeout(timeout_ms, stream.close()).await {
Ok(Ok(())) => {}
Ok(out) Ok(Err(e)) => {
return Ok(err_to_network_result(e));
// Drive connection to close
/*
let cur_ts = get_timestamp();
loop {
match stream.flush().await {
Ok(()) => {}
Err(Error::Io(ioerr)) => {
break Err(ioerr).into_network_result();
}
Err(Error::ConnectionClosed) => {
break Ok(NetworkResult::value(()));
}
Err(e) => {
break Err(to_io_error_other(e));
}
} }
if get_timestamp().saturating_sub(cur_ts) >= MAX_CONNECTION_CLOSE_WAIT_US { Err(_) => {
return Ok(NetworkResult::Timeout); // Timed out
return Ok(NetworkResult::timeout());
} }
} };
*/
Ok(NetworkResult::value(()))
} }
#[instrument(level = "trace", target="protocol", err, skip(self, message), fields(network_result, message.len = message.len()))] #[instrument(level = "trace", target="protocol", err, skip(self, message), fields(network_result, message.len = message.len()))]

View file

@ -480,20 +480,21 @@ impl NetworkConnection {
} }
} }
veilid_log!(registry trace
"Connection loop finished flow={:?}",
flow
);
// Let the connection manager know the receive loop exited // Let the connection manager know the receive loop exited
connection_manager connection_manager
.report_connection_finished(connection_id) .report_connection_finished(connection_id);
.await;
// Close the low level socket // Close the low level socket
if let Err(e) = protocol_connection.close().await { if let Err(e) = protocol_connection.close().await {
veilid_log!(registry debug "Protocol connection close error: {}", e); veilid_log!(registry debug "Protocol connection close error: {}", e);
} }
veilid_log!(registry trace
"Connection loop exited flow={:?}",
flow
);
}.in_current_span()) }.in_current_span())
} }

View file

@ -152,7 +152,7 @@ impl NetworkManager {
// If a node is unreachable it may still have an existing inbound connection // If a node is unreachable it may still have an existing inbound connection
// Try that, but don't cache anything // Try that, but don't cache anything
network_result_try!( network_result_try!(
pin_future_closure!(self.send_data_ncm_existing(target_node_ref, data)).await? pin_future_closure!(self.send_data_unreachable(target_node_ref, data)).await?
) )
} }
Some(NodeContactMethod { Some(NodeContactMethod {
@ -239,6 +239,42 @@ impl NetworkManager {
})) }))
} }
/// Send data to unreachable node
#[instrument(level = "trace", target = "net", skip_all, err)]
async fn send_data_unreachable(
&self,
target_node_ref: FilteredNodeRef,
data: Vec<u8>,
) -> EyreResult<NetworkResult<UniqueFlow>> {
// First try to send data to the last connection we've seen this peer on
let Some(flow) = target_node_ref.last_flow() else {
return Ok(NetworkResult::no_connection_other(format!(
"node was unreachable: {}",
target_node_ref
)));
};
let net = self.net();
let unique_flow = match pin_future!(debug_duration(
|| { net.send_data_to_existing_flow(flow, data) },
Some(1_000_000)
))
.await?
{
SendDataToExistingFlowResult::Sent(unique_flow) => unique_flow,
SendDataToExistingFlowResult::NotSent(_) => {
return Ok(NetworkResult::no_connection_other(
"failed to send to existing flow",
));
}
};
// Update timestamp for this last connection since we just sent to it
self.set_last_flow(target_node_ref.unfiltered(), flow, Timestamp::now());
Ok(NetworkResult::value(unique_flow))
}
/// Send data using NodeContactMethod::Existing /// Send data using NodeContactMethod::Existing
#[instrument(level = "trace", target = "net", skip_all, err)] #[instrument(level = "trace", target = "net", skip_all, err)]
async fn send_data_ncm_existing( async fn send_data_ncm_existing(
@ -255,7 +291,12 @@ impl NetworkManager {
}; };
let net = self.net(); let net = self.net();
let unique_flow = match pin_future!(net.send_data_to_existing_flow(flow, data)).await? { let unique_flow = match pin_future!(debug_duration(
|| { net.send_data_to_existing_flow(flow, data) },
Some(1_000_000)
))
.await?
{
SendDataToExistingFlowResult::Sent(unique_flow) => unique_flow, SendDataToExistingFlowResult::Sent(unique_flow) => unique_flow,
SendDataToExistingFlowResult::NotSent(_) => { SendDataToExistingFlowResult::NotSent(_) => {
return Ok(NetworkResult::no_connection_other( return Ok(NetworkResult::no_connection_other(
@ -297,7 +338,12 @@ impl NetworkManager {
// First try to send data to the last flow we've seen this peer on // First try to send data to the last flow we've seen this peer on
let data = if let Some(flow) = seq_target_node_ref.last_flow() { let data = if let Some(flow) = seq_target_node_ref.last_flow() {
let net = self.net(); let net = self.net();
match pin_future!(net.send_data_to_existing_flow(flow, data)).await? { match pin_future!(debug_duration(
|| { net.send_data_to_existing_flow(flow, data) },
Some(1_000_000)
))
.await?
{
SendDataToExistingFlowResult::Sent(unique_flow) => { SendDataToExistingFlowResult::Sent(unique_flow) => {
// Update timestamp for this last connection since we just sent to it // Update timestamp for this last connection since we just sent to it
self.set_last_flow(target_node_ref.unfiltered(), flow, Timestamp::now()); self.set_last_flow(target_node_ref.unfiltered(), flow, Timestamp::now());
@ -321,9 +367,16 @@ impl NetworkManager {
data data
}; };
let connection_initial_timeout_us = self
.config()
.with(|c| c.network.connection_initial_timeout_ms as u64 * 1000);
let unique_flow = network_result_try!( let unique_flow = network_result_try!(
pin_future!(self.do_reverse_connect(relay_nr.clone(), target_node_ref.clone(), data)) pin_future!(debug_duration(
.await? || { self.do_reverse_connect(relay_nr.clone(), target_node_ref.clone(), data) },
Some(connection_initial_timeout_us * 2)
))
.await?
); );
Ok(NetworkResult::value(unique_flow)) Ok(NetworkResult::value(unique_flow))
} }
@ -339,7 +392,12 @@ impl NetworkManager {
// First try to send data to the last flow we've seen this peer on // First try to send data to the last flow we've seen this peer on
let data = if let Some(flow) = target_node_ref.last_flow() { let data = if let Some(flow) = target_node_ref.last_flow() {
let net = self.net(); let net = self.net();
match pin_future!(net.send_data_to_existing_flow(flow, data)).await? { match pin_future!(debug_duration(
|| { net.send_data_to_existing_flow(flow, data) },
Some(1_000_000)
))
.await?
{
SendDataToExistingFlowResult::Sent(unique_flow) => { SendDataToExistingFlowResult::Sent(unique_flow) => {
// Update timestamp for this last connection since we just sent to it // Update timestamp for this last connection since we just sent to it
self.set_last_flow(target_node_ref.unfiltered(), flow, Timestamp::now()); self.set_last_flow(target_node_ref.unfiltered(), flow, Timestamp::now());
@ -363,9 +421,16 @@ impl NetworkManager {
data data
}; };
let hole_punch_receipt_time_us = self
.config()
.with(|c| c.network.hole_punch_receipt_time_ms as u64 * 1000);
let unique_flow = network_result_try!( let unique_flow = network_result_try!(
pin_future!(self.do_hole_punch(relay_nr.clone(), target_node_ref.clone(), data)) pin_future!(debug_duration(
.await? || { self.do_hole_punch(relay_nr.clone(), target_node_ref.clone(), data) },
Some(hole_punch_receipt_time_us * 2)
))
.await?
); );
Ok(NetworkResult::value(unique_flow)) Ok(NetworkResult::value(unique_flow))
@ -391,7 +456,12 @@ impl NetworkManager {
); );
let net = self.net(); let net = self.net();
match pin_future!(net.send_data_to_existing_flow(flow, data)).await? { match pin_future!(debug_duration(
|| { net.send_data_to_existing_flow(flow, data) },
Some(1_000_000)
))
.await?
{
SendDataToExistingFlowResult::Sent(unique_flow) => { SendDataToExistingFlowResult::Sent(unique_flow) => {
// Update timestamp for this last connection since we just sent to it // Update timestamp for this last connection since we just sent to it
self.set_last_flow(node_ref.unfiltered(), flow, Timestamp::now()); self.set_last_flow(node_ref.unfiltered(), flow, Timestamp::now());

View file

@ -52,8 +52,16 @@ impl WebsocketNetworkConnection {
instrument(level = "trace", err, skip(self)) instrument(level = "trace", err, skip(self))
)] )]
pub async fn close(&self) -> io::Result<NetworkResult<()>> { pub async fn close(&self) -> io::Result<NetworkResult<()>> {
let timeout_ms = self
.registry
.config()
.with(|c| c.network.connection_initial_timeout_ms);
#[allow(unused_variables)] #[allow(unused_variables)]
let x = self.inner.ws_meta.close().await.map_err(ws_err_to_io_error); let x = match timeout(timeout_ms, self.inner.ws_meta.close()).await {
Ok(v) => v.map_err(ws_err_to_io_error),
Err(_) => return Ok(NetworkResult::timeout()),
};
#[cfg(feature = "verbose-tracing")] #[cfg(feature = "verbose-tracing")]
veilid_log!(self debug "close result: {:?}", x); veilid_log!(self debug "close result: {:?}", x);
Ok(NetworkResult::value(())) Ok(NetworkResult::value(()))

View file

@ -1011,7 +1011,12 @@ impl BucketEntryInner {
match latest_contact_time { match latest_contact_time {
None => { None => {
error!("Peer is reliable, but not seen!"); // Peer may be appear reliable from a previous attach/detach
// But reliability uses last_seen_ts not the last_outbound_contact_time
// Regardless, if we haven't pinged it, we need to ping it.
// But it it was reliable before, and pings successfully then it can
// stay reliable, so we don't make it unreliable just because we haven't
// contacted it yet during this attachment.
true true
} }
Some(latest_contact_time) => { Some(latest_contact_time) => {

View file

@ -1041,7 +1041,7 @@ impl RoutingTable {
#[instrument(level = "trace", skip(self), err)] #[instrument(level = "trace", skip(self), err)]
pub async fn find_nodes_close_to_node_id( pub async fn find_nodes_close_to_node_id(
&self, &self,
node_ref: NodeRef, node_ref: FilteredNodeRef,
node_id: TypedKey, node_id: TypedKey,
capabilities: Vec<Capability>, capabilities: Vec<Capability>,
) -> EyreResult<NetworkResult<Vec<NodeRef>>> { ) -> EyreResult<NetworkResult<Vec<NodeRef>>> {
@ -1049,11 +1049,7 @@ impl RoutingTable {
let res = network_result_try!( let res = network_result_try!(
rpc_processor rpc_processor
.rpc_call_find_node( .rpc_call_find_node(Destination::direct(node_ref), node_id, capabilities)
Destination::direct(node_ref.default_filtered()),
node_id,
capabilities
)
.await? .await?
); );
@ -1069,7 +1065,7 @@ impl RoutingTable {
pub async fn find_nodes_close_to_self( pub async fn find_nodes_close_to_self(
&self, &self,
crypto_kind: CryptoKind, crypto_kind: CryptoKind,
node_ref: NodeRef, node_ref: FilteredNodeRef,
capabilities: Vec<Capability>, capabilities: Vec<Capability>,
) -> EyreResult<NetworkResult<Vec<NodeRef>>> { ) -> EyreResult<NetworkResult<Vec<NodeRef>>> {
let self_node_id = self.node_id(crypto_kind); let self_node_id = self.node_id(crypto_kind);
@ -1083,7 +1079,7 @@ impl RoutingTable {
pub async fn find_nodes_close_to_node_ref( pub async fn find_nodes_close_to_node_ref(
&self, &self,
crypto_kind: CryptoKind, crypto_kind: CryptoKind,
node_ref: NodeRef, node_ref: FilteredNodeRef,
capabilities: Vec<Capability>, capabilities: Vec<Capability>,
) -> EyreResult<NetworkResult<Vec<NodeRef>>> { ) -> EyreResult<NetworkResult<Vec<NodeRef>>> {
let Some(target_node_id) = node_ref.node_ids().get(crypto_kind) else { let Some(target_node_id) = node_ref.node_ids().get(crypto_kind) else {
@ -1104,7 +1100,7 @@ impl RoutingTable {
capabilities: Vec<Capability>, capabilities: Vec<Capability>,
) { ) {
// Ask node for nodes closest to our own node // Ask node for nodes closest to our own node
let closest_nodes = network_result_value_or_log!(self match pin_future!(self.find_nodes_close_to_self(crypto_kind, node_ref.clone(), capabilities.clone())).await { let closest_nodes = network_result_value_or_log!(self match pin_future!(self.find_nodes_close_to_self(crypto_kind, node_ref.sequencing_filtered(Sequencing::PreferOrdered), capabilities.clone())).await {
Err(e) => { Err(e) => {
veilid_log!(self error veilid_log!(self error
"find_self failed for {:?}: {:?}", "find_self failed for {:?}: {:?}",
@ -1120,7 +1116,7 @@ impl RoutingTable {
// Ask each node near us to find us as well // Ask each node near us to find us as well
if wide { if wide {
for closest_nr in closest_nodes { for closest_nr in closest_nodes {
network_result_value_or_log!(self match pin_future!(self.find_nodes_close_to_self(crypto_kind, closest_nr.clone(), capabilities.clone())).await { network_result_value_or_log!(self match pin_future!(self.find_nodes_close_to_self(crypto_kind, closest_nr.sequencing_filtered(Sequencing::PreferOrdered), capabilities.clone())).await {
Err(e) => { Err(e) => {
veilid_log!(self error veilid_log!(self error
"find_self failed for {:?}: {:?}", "find_self failed for {:?}: {:?}",

View file

@ -119,6 +119,24 @@ impl NodeRefOperateTrait for FilteredNodeRef {
let inner = &mut *routing_table.inner.write(); let inner = &mut *routing_table.inner.write();
self.entry.with_mut(inner, f) self.entry.with_mut(inner, f)
} }
fn with_inner<T, F>(&self, f: F) -> T
where
F: FnOnce(&RoutingTableInner) -> T,
{
let routing_table = self.registry.routing_table();
let inner = &*routing_table.inner.read();
f(inner)
}
fn with_inner_mut<T, F>(&self, f: F) -> T
where
F: FnOnce(&mut RoutingTableInner) -> T,
{
let routing_table = self.registry.routing_table();
let inner = &mut *routing_table.inner.write();
f(inner)
}
} }
impl NodeRefCommonTrait for FilteredNodeRef {} impl NodeRefCommonTrait for FilteredNodeRef {}

View file

@ -139,6 +139,24 @@ impl NodeRefOperateTrait for NodeRef {
let inner = &mut *routing_table.inner.write(); let inner = &mut *routing_table.inner.write();
self.entry.with_mut(inner, f) self.entry.with_mut(inner, f)
} }
fn with_inner<T, F>(&self, f: F) -> T
where
F: FnOnce(&RoutingTableInner) -> T,
{
let routing_table = self.routing_table();
let inner = &*routing_table.inner.read();
f(inner)
}
fn with_inner_mut<T, F>(&self, f: F) -> T
where
F: FnOnce(&mut RoutingTableInner) -> T,
{
let routing_table = self.routing_table();
let inner = &mut *routing_table.inner.write();
f(inner)
}
} }
impl NodeRefCommonTrait for NodeRef {} impl NodeRefCommonTrait for NodeRef {}

View file

@ -90,6 +90,21 @@ impl<'a, N: NodeRefAccessorsTrait + NodeRefOperateTrait + fmt::Debug + fmt::Disp
{ {
panic!("need to locked_mut() for this operation") panic!("need to locked_mut() for this operation")
} }
fn with_inner<T, F>(&self, f: F) -> T
where
F: FnOnce(&RoutingTableInner) -> T,
{
let inner = &*self.inner.lock();
f(inner)
}
fn with_inner_mut<T, F>(&self, _f: F) -> T
where
F: FnOnce(&mut RoutingTableInner) -> T,
{
panic!("need to locked_mut() for this operation")
}
} }
impl<'a, N: NodeRefAccessorsTrait + NodeRefOperateTrait + fmt::Debug + fmt::Display + Clone> impl<'a, N: NodeRefAccessorsTrait + NodeRefOperateTrait + fmt::Debug + fmt::Display + Clone>

View file

@ -92,6 +92,22 @@ impl<'a, N: NodeRefAccessorsTrait + NodeRefOperateTrait + fmt::Debug + fmt::Disp
let inner = &mut *self.inner.lock(); let inner = &mut *self.inner.lock();
self.nr.entry().with_mut(inner, f) self.nr.entry().with_mut(inner, f)
} }
fn with_inner<T, F>(&self, f: F) -> T
where
F: FnOnce(&RoutingTableInner) -> T,
{
let inner = &*self.inner.lock();
f(inner)
}
fn with_inner_mut<T, F>(&self, f: F) -> T
where
F: FnOnce(&mut RoutingTableInner) -> T,
{
let inner = &mut *self.inner.lock();
f(inner)
}
} }
impl<'a, N: NodeRefAccessorsTrait + NodeRefOperateTrait + fmt::Debug + fmt::Display + Clone> impl<'a, N: NodeRefAccessorsTrait + NodeRefOperateTrait + fmt::Debug + fmt::Display + Clone>

View file

@ -20,6 +20,13 @@ pub(crate) trait NodeRefOperateTrait {
fn operate_mut<T, F>(&self, f: F) -> T fn operate_mut<T, F>(&self, f: F) -> T
where where
F: FnOnce(&mut RoutingTableInner, &mut BucketEntryInner) -> T; F: FnOnce(&mut RoutingTableInner, &mut BucketEntryInner) -> T;
#[expect(dead_code)]
fn with_inner<T, F>(&self, f: F) -> T
where
F: FnOnce(&RoutingTableInner) -> T;
fn with_inner_mut<T, F>(&self, f: F) -> T
where
F: FnOnce(&mut RoutingTableInner) -> T;
} }
// Common Operations // Common Operations
@ -115,7 +122,7 @@ pub(crate) trait NodeRefCommonTrait: NodeRefAccessorsTrait + NodeRefOperateTrait
// } // }
fn relay(&self, routing_domain: RoutingDomain) -> EyreResult<Option<FilteredNodeRef>> { fn relay(&self, routing_domain: RoutingDomain) -> EyreResult<Option<FilteredNodeRef>> {
self.operate_mut(|rti, e| { let Some(rpi) = self.operate(|rti, e| {
let Some(sni) = e.signed_node_info(routing_domain) else { let Some(sni) = e.signed_node_info(routing_domain) else {
return Ok(None); return Ok(None);
}; };
@ -127,8 +134,14 @@ pub(crate) trait NodeRefCommonTrait: NodeRefAccessorsTrait + NodeRefOperateTrait
if rti.routing_table().matches_own_node_id(rpi.node_ids()) { if rti.routing_table().matches_own_node_id(rpi.node_ids()) {
bail!("Can't relay though ourselves"); bail!("Can't relay though ourselves");
} }
Ok(Some(rpi))
})?
else {
return Ok(None);
};
// Register relay node and return noderef // Register relay node and return noderef
self.with_inner_mut(|rti| {
let nr = rti.register_node_with_peer_info(rpi, false)?; let nr = rti.register_node_with_peer_info(rpi, false)?;
Ok(Some(nr)) Ok(Some(nr))
}) })

View file

@ -716,7 +716,7 @@ impl RouteSpecStore {
}; };
let Some(rsid) = inner.content.get_id_by_key(&public_key.value) else { let Some(rsid) = inner.content.get_id_by_key(&public_key.value) else {
veilid_log!(self debug "route id does not exist: {:?}", public_key.value); veilid_log!(self debug target: "network_result", "route id does not exist: {:?}", public_key.value);
return None; return None;
}; };
let Some(rssd) = inner.content.get_detail(&rsid) else { let Some(rssd) = inner.content.get_detail(&rsid) else {
@ -753,7 +753,7 @@ impl RouteSpecStore {
return None; return None;
} }
Err(e) => { Err(e) => {
veilid_log!(self debug "errir verifying signature for hop {} at {} on private route {}: {}", hop_n, hop_public_key, public_key, e); veilid_log!(self debug "error verifying signature for hop {} at {} on private route {}: {}", hop_n, hop_public_key, public_key, e);
return None; return None;
} }
} }

View file

@ -289,7 +289,7 @@ impl RoutingTable {
// Get what contact method would be used for contacting the bootstrap // Get what contact method would be used for contacting the bootstrap
let bsdi = match network_manager let bsdi = match network_manager
.get_node_contact_method(nr.default_filtered()) .get_node_contact_method(nr.sequencing_filtered(Sequencing::PreferOrdered))
{ {
Ok(Some(ncm)) if ncm.is_direct() => ncm.direct_dial_info().unwrap(), Ok(Some(ncm)) if ncm.is_direct() => ncm.direct_dial_info().unwrap(),
Ok(v) => { Ok(v) => {
@ -307,7 +307,7 @@ impl RoutingTable {
// Need VALID signed peer info, so ask bootstrap to find_node of itself // Need VALID signed peer info, so ask bootstrap to find_node of itself
// which will ensure it has the bootstrap's signed peer info as part of the response // which will ensure it has the bootstrap's signed peer info as part of the response
let _ = routing_table.find_nodes_close_to_node_ref(crypto_kind, nr.clone(), vec![]).await; let _ = routing_table.find_nodes_close_to_node_ref(crypto_kind, nr.sequencing_filtered(Sequencing::PreferOrdered), vec![]).await;
// Ensure we got the signed peer info // Ensure we got the signed peer info
if !nr.signed_node_info_has_valid_signature(routing_domain) { if !nr.signed_node_info_has_valid_signature(routing_domain) {

View file

@ -1471,11 +1471,24 @@ impl RPCProcessor {
let operation = match self.decode_rpc_operation(&encoded_msg) { let operation = match self.decode_rpc_operation(&encoded_msg) {
Ok(v) => v, Ok(v) => v,
Err(e) => { Err(e) => {
// Debug on error match e {
veilid_log!(self debug "Dropping routed RPC: {}", e); // Invalid messages that should be punished
RPCError::Protocol(_) | RPCError::InvalidFormat(_) => {
veilid_log!(self debug "Invalid routed RPC Operation: {}", e);
// XXX: Punish routes that send routed undecodable crap
// self.network_manager().address_filter().punish_route_id(xxx, PunishmentReason::FailedToDecodeRoutedMessage);
}
// Ignored messages that should be dropped
RPCError::Ignore(_) | RPCError::Network(_) | RPCError::TryAgain(_) => {
veilid_log!(self trace "Dropping routed RPC Operation: {}", e);
}
// Internal errors that deserve louder logging
RPCError::Unimplemented(_) | RPCError::Internal(_) => {
veilid_log!(self error "Error decoding routed RPC operation: {}", e);
}
};
// XXX: Punish routes that send routed undecodable crap
// self.network_manager().address_filter().punish_route_id(xxx, PunishmentReason::FailedToDecodeRoutedMessage);
return Ok(NetworkResult::invalid_message(e)); return Ok(NetworkResult::invalid_message(e));
} }
}; };
@ -1593,16 +1606,16 @@ impl RPCProcessor {
if let Err(e) = self.waiting_rpc_table.complete_op_waiter(op_id, msg) { if let Err(e) = self.waiting_rpc_table.complete_op_waiter(op_id, msg) {
match e { match e {
RPCError::Unimplemented(_) | RPCError::Internal(_) => { RPCError::Unimplemented(_) | RPCError::Internal(_) => {
veilid_log!(self error "Could not complete rpc operation: id = {}: {}", op_id, e); veilid_log!(self error "Error in RPC operation: id = {}: {}", op_id, e);
} }
RPCError::InvalidFormat(_) RPCError::InvalidFormat(_)
| RPCError::Protocol(_) | RPCError::Protocol(_)
| RPCError::Network(_) | RPCError::Network(_)
| RPCError::TryAgain(_) => { | RPCError::TryAgain(_) => {
veilid_log!(self debug "Could not complete rpc operation: id = {}: {}", op_id, e); veilid_log!(self debug "Could not complete RPC operation: id = {}: {}", op_id, e);
} }
RPCError::Ignore(_) => { RPCError::Ignore(e) => {
veilid_log!(self debug "Answer late: id = {}", op_id); veilid_log!(self debug "RPC operation ignored: id = {}: {}", op_id, e);
} }
}; };
// Don't throw an error here because it's okay if the original operation timed out // Don't throw an error here because it's okay if the original operation timed out

View file

@ -8,7 +8,7 @@ where
{ {
waiter: OperationWaiter<T, C>, waiter: OperationWaiter<T, C>,
op_id: OperationId, op_id: OperationId,
result_receiver: Option<flume::Receiver<(Span, T)>>, result_receiver: flume::Receiver<(Span, T)>,
} }
impl<T, C> OperationWaitHandle<T, C> impl<T, C> OperationWaitHandle<T, C>
@ -27,9 +27,7 @@ where
C: Unpin + Clone, C: Unpin + Clone,
{ {
fn drop(&mut self) { fn drop(&mut self) {
if self.result_receiver.is_some() { self.waiter.cancel_op_waiter(self.op_id);
self.waiter.cancel_op_waiter(self.op_id);
}
} }
} }
@ -106,7 +104,7 @@ where
OperationWaitHandle { OperationWaitHandle {
waiter: self.clone(), waiter: self.clone(),
op_id, op_id,
result_receiver: Some(result_receiver), result_receiver,
} }
} }
@ -125,65 +123,69 @@ where
/// Get operation context /// Get operation context
pub fn get_op_context(&self, op_id: OperationId) -> Result<C, RPCError> { pub fn get_op_context(&self, op_id: OperationId) -> Result<C, RPCError> {
let inner = self.inner.lock(); let inner = self.inner.lock();
let Some(waiting_op) = inner.waiting_op_table.get(&op_id) else { let res = {
return Err(RPCError::ignore(format!( let Some(waiting_op) = inner.waiting_op_table.get(&op_id) else {
"Missing operation id getting op context: id={}", return Err(RPCError::ignore(format!(
op_id "Missing operation id getting op context: id={}",
))); op_id
)));
};
Ok(waiting_op.context.clone())
}; };
Ok(waiting_op.context.clone()) drop(inner);
res
} }
/// Remove wait for op /// Remove wait for op
#[instrument(level = "trace", target = "rpc", skip_all)] #[instrument(level = "trace", target = "rpc", skip_all)]
fn cancel_op_waiter(&self, op_id: OperationId) { fn cancel_op_waiter(&self, op_id: OperationId) {
let mut inner = self.inner.lock(); let mut inner = self.inner.lock();
inner.waiting_op_table.remove(&op_id); {
let waiting_op = inner.waiting_op_table.remove(&op_id);
drop(waiting_op);
}
drop(inner);
} }
/// Complete the waiting op /// Complete the waiting op
#[instrument(level = "trace", target = "rpc", skip_all)] #[instrument(level = "trace", target = "rpc", skip_all)]
pub fn complete_op_waiter(&self, op_id: OperationId, message: T) -> Result<(), RPCError> { pub fn complete_op_waiter(&self, op_id: OperationId, message: T) -> Result<(), RPCError> {
let waiting_op = { let mut inner = self.inner.lock();
let mut inner = self.inner.lock(); let res = {
inner let waiting_op =
.waiting_op_table inner
.remove(&op_id) .waiting_op_table
.ok_or_else(RPCError::else_ignore(format!( .remove(&op_id)
"Unmatched operation id: {}", .ok_or_else(RPCError::else_ignore(format!(
op_id "Unmatched operation id: {}",
)))? op_id
)))?;
waiting_op
.result_sender
.send((Span::current(), message))
.map_err(RPCError::ignore)
}; };
waiting_op drop(inner);
.result_sender res
.send((Span::current(), message))
.map_err(RPCError::ignore)
} }
/// Wait for operation to complete /// Wait for operation to complete
#[instrument(level = "trace", target = "rpc", skip_all)] #[instrument(level = "trace", target = "rpc", skip_all)]
pub async fn wait_for_op( pub async fn wait_for_op(
&self, &self,
mut handle: OperationWaitHandle<T, C>, handle: OperationWaitHandle<T, C>,
timeout_us: TimestampDuration, timeout_us: TimestampDuration,
) -> Result<TimeoutOr<(T, TimestampDuration)>, RPCError> { ) -> Result<TimeoutOr<(T, TimestampDuration)>, RPCError> {
let timeout_ms = us_to_ms(timeout_us.as_u64()).map_err(RPCError::internal)?; let timeout_ms = us_to_ms(timeout_us.as_u64()).map_err(RPCError::internal)?;
// Take the receiver let result_fut = handle.result_receiver.recv_async().in_current_span();
// After this, we must manually cancel since the cancel on handle drop is disabled
let result_receiver = handle.result_receiver.take().unwrap();
let result_fut = result_receiver.recv_async().in_current_span();
// wait for eventualvalue // wait for eventualvalue
let start_ts = Timestamp::now(); let start_ts = Timestamp::now();
let res = timeout(timeout_ms, result_fut).await.into_timeout_or(); let res = timeout(timeout_ms, result_fut).await.into_timeout_or();
match res { match res {
TimeoutOr::Timeout => { TimeoutOr::Timeout => Ok(TimeoutOr::Timeout),
self.cancel_op_waiter(handle.op_id);
Ok(TimeoutOr::Timeout)
}
TimeoutOr::Value(Ok((_span_id, ret))) => { TimeoutOr::Value(Ok((_span_id, ret))) => {
let end_ts = Timestamp::now(); let end_ts = Timestamp::now();
@ -192,7 +194,10 @@ where
Ok(TimeoutOr::Value((ret, end_ts.saturating_sub(start_ts)))) Ok(TimeoutOr::Value((ret, end_ts.saturating_sub(start_ts))))
} }
TimeoutOr::Value(Err(e)) => Err(RPCError::ignore(e)), TimeoutOr::Value(Err(e)) => {
//
Err(RPCError::ignore(e))
}
} }
} }
} }

View file

@ -70,7 +70,7 @@ impl RPCProcessor {
match request.kind { match request.kind {
// Process RPC Message // Process RPC Message
RPCWorkerRequestKind::Message { message_encoded } => { RPCWorkerRequestKind::Message { message_encoded } => {
network_result_value_or_log!(self match self network_result_value_or_log!(self target:"network_result", match self
.process_rpc_message(message_encoded).instrument(rpc_request_span) .process_rpc_message(message_encoded).instrument(rpc_request_span)
.await .await
{ {

View file

@ -1120,7 +1120,7 @@ impl StorageManager {
let dest = rpc_processor let dest = rpc_processor
.resolve_target_to_destination( .resolve_target_to_destination(
vc.target, vc.target,
SafetySelection::Unsafe(Sequencing::NoPreference), SafetySelection::Unsafe(Sequencing::PreferOrdered),
) )
.await .await
.map_err(VeilidAPIError::from)?; .map_err(VeilidAPIError::from)?;

View file

@ -47,6 +47,7 @@ veilid_tools_android_tests = ["dep:paranoid-android"]
veilid_tools_ios_tests = ["dep:tracing", "dep:oslog", "dep:tracing-oslog"] veilid_tools_ios_tests = ["dep:tracing", "dep:oslog", "dep:tracing-oslog"]
tracing = ["dep:tracing", "dep:tracing-subscriber", "tokio/tracing"] tracing = ["dep:tracing", "dep:tracing-subscriber", "tokio/tracing"]
debug-locks = [] debug-locks = []
debug-duration-timeout = []
virtual-network = [] virtual-network = []
virtual-network-server = [ virtual-network-server = [

View file

@ -10,7 +10,7 @@ where
{ {
table: AsyncTagLockTable<T>, table: AsyncTagLockTable<T>,
tag: T, tag: T,
_guard: AsyncMutexGuardArc<()>, guard: Option<AsyncMutexGuardArc<()>>,
} }
impl<T> AsyncTagLockGuard<T> impl<T> AsyncTagLockGuard<T>
@ -21,7 +21,7 @@ where
Self { Self {
table, table,
tag, tag,
_guard: guard, guard: Some(guard),
} }
} }
} }
@ -45,7 +45,8 @@ where
if guards == 0 { if guards == 0 {
inner.table.remove(&self.tag).unwrap(); inner.table.remove(&self.tag).unwrap();
} }
// Proceed with releasing _guard, which may cause some concurrent tag lock to acquire // Proceed with releasing guard, which may cause some concurrent tag lock to acquire
drop(self.guard.take());
} }
} }
@ -153,7 +154,7 @@ where
} }
std::collections::hash_map::Entry::Vacant(v) => { std::collections::hash_map::Entry::Vacant(v) => {
let mutex = Arc::new(AsyncMutex::new(())); let mutex = Arc::new(AsyncMutex::new(()));
let guard = asyncmutex_try_lock_arc!(mutex)?; let guard = asyncmutex_try_lock_arc!(mutex).unwrap();
v.insert(AsyncTagLockTableEntry { mutex, guards: 1 }); v.insert(AsyncTagLockTableEntry { mutex, guards: 1 });
guard guard
} }

View file

@ -278,85 +278,3 @@ macro_rules! network_result_try {
} }
}; };
} }
#[macro_export]
macro_rules! network_result_value_or_log {
($self:ident $r:expr => $f:expr) => {
network_result_value_or_log!($self $r => [ "" ] $f )
};
($self:ident $r:expr => [ $d:expr ] $f:expr) => { {
let __extra_message = if debug_target_enabled!("network_result") {
$d.to_string()
} else {
"".to_string()
};
match $r {
NetworkResult::Timeout => {
veilid_log!($self debug
"{} at {}@{}:{} in {}{}",
"Timeout",
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::ServiceUnavailable(ref s) => {
veilid_log!($self debug
"{}({}) at {}@{}:{} in {}{}",
"ServiceUnavailable",
s,
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::NoConnection(ref e) => {
veilid_log!($self debug
"{}({}) at {}@{}:{} in {}{}",
"No connection",
e.to_string(),
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::AlreadyExists(ref e) => {
veilid_log!($self debug
"{}({}) at {}@{}:{} in {}{}",
"Already exists",
e.to_string(),
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::InvalidMessage(ref s) => {
veilid_log!($self debug
"{}({}) at {}@{}:{} in {}{}",
"Invalid message",
s,
file!(),
line!(),
column!(),
fn_name::uninstantiated!(),
__extra_message
);
$f
}
NetworkResult::Value(v) => v,
}
} };
}

View file

@ -522,13 +522,33 @@ pub fn is_debug_backtrace_enabled() -> bool {
} }
#[track_caller] #[track_caller]
pub fn debug_duration<R, F: Future<Output = R>, T: FnOnce() -> F>(f: T) -> impl Future<Output = R> { pub fn debug_duration<R, F: Future<Output = R>, T: FnOnce() -> F>(
let location = std::panic::Location::caller(); f: T,
opt_timeout_us: Option<u64>,
) -> impl Future<Output = R> {
let location = core::panic::Location::caller();
async move { async move {
let t1 = get_timestamp(); let t1 = get_timestamp();
let out = f().await; let out = f().await;
let t2 = get_timestamp(); let t2 = get_timestamp();
debug!("duration@{}: {}", location, display_duration(t2 - t1)); let duration_us = t2 - t1;
if let Some(timeout_us) = opt_timeout_us {
if duration_us > timeout_us {
#[cfg(not(feature = "debug-duration-timeout"))]
debug!(
"Excessive duration: {}\n{:?}",
display_duration(duration_us),
backtrace::Backtrace::new()
);
#[cfg(feature = "debug-duration-timeout")]
panic!(format!(
"Duration panic timeout exceeded: {}",
display_duration(duration_us)
));
}
} else {
debug!("Duration: {} = {}", location, display_duration(duration_us),);
}
out out
} }
} }