maybe maybe

## Summary
- When a node times out, its info gets cleared from state. The monitor
functions only sent data when something changed, leaving no mechanism to
re-populate this info after a timeout.
- Removes change-detection checks from `_monitor_misc`,
`_monitor_system_profiler_thunderbolt_data`, `_watch_system_info`, and
`_monitor_thunderbolt_bridge_status` so data is sent periodically
regardless of whether it changed.

## Test plan
- [ ] Verify type checker passes: `uv run basedpyright`
- [ ] Verify linter passes: `uv run ruff check`
- [ ] Verify tests pass: `uv run pytest`
- [ ] Manually test that node info is re-populated after a timeout by
observing cluster behavior

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Cargo.lock

@@ -514,6 +514,20 @@ version = "0.7.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a1d728cc89cf3aee9ff92b05e62b19ee65a02b5702cff7d5a377e32c6ae29d8d"
 
+[[package]]
+name = "cluster_membership"
+version = "0.0.1"
+dependencies = [
+ "anyhow",
+ "async-trait",
+ "futures-lite",
+ "futures-timer",
+ "libp2p",
+ "log",
+ "tokio",
+ "tracing-subscriber",
+]
+
 [[package]]
 name = "colorchoice"
 version = "1.0.4"
@@ -998,6 +1012,7 @@ dependencies = [
 name = "exo_pyo3_bindings"
 version = "0.0.1"
 dependencies = [
+ "cluster_membership",
  "delegate",
  "derive_more",
  "env_logger",
@@ -1030,6 +1045,12 @@ dependencies = [
  "syn 2.0.111",
 ]
 
+[[package]]
+name = "fastrand"
+version = "2.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
+
 [[package]]
 name = "ff"
 version = "0.13.1"
@@ -1138,7 +1159,10 @@ version = "2.6.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f78e10609fe0e0b3f4157ffab1876319b5b0db102a2c60dc4626306dc46b44ad"
 dependencies = [
+ "fastrand",
  "futures-core",
+ "futures-io",
+ "parking",
  "pin-project-lite",
 ]
 

Cargo.toml

@@ -4,6 +4,7 @@ members = [
     "rust/networking",
     "rust/exo_pyo3_bindings",
     "rust/util",
+    "rust/cluster_membership",
 ]
 
 [workspace.package]
@@ -25,6 +26,7 @@ opt-level = 3
 ## Crate members as common dependencies
 networking = { path = "rust/networking" }
 util = { path = "rust/util" }
+cluster_membership = { path = "rust/cluster_membership" }
 
 # Proc-macro authoring tools
 syn = "2.0"
@@ -62,6 +64,7 @@ frunk-enum-core = "0.3"
 # Async dependencies
 tokio = "1.46"
 futures = "0.3"
+futures-lite = "2.6.1"
 futures-util = "0.3"
 futures-timer = "3.0"
 

app/EXO/EXO/Services/NetworkSetupHelper.swift

@@ -31,6 +31,35 @@ enum NetworkSetupHelper {
         # Remove Thunderbolt Bridge from VirtualNetworkInterfaces in preferences.plist
         /usr/libexec/PlistBuddy -c "Delete :VirtualNetworkInterfaces:Bridge:bridge0" "$PREFS" 2>/dev/null || true
 
+        networksetup -listlocations | grep -q exo || {
+          networksetup -createlocation exo
+        }
+
+        networksetup -switchtolocation exo
+        networksetup -listallhardwareports \\
+          | awk -F': ' '/Hardware Port: / {print $2}' \\
+          | while IFS=":" read -r name; do
+              case "$name" in
+                "Ethernet Adapter"*)
+                        ;;
+                "Thunderbolt Bridge")
+                        ;;
+                "Thunderbolt "*)
+                  networksetup -listallnetworkservices \\
+                    | grep -q "EXO $name" \\
+                      || networksetup -createnetworkservice "EXO $name" "$name" 2>/dev/null \\
+                      || continue
+                  networksetup -setdhcp "EXO $name"
+                        ;;
+                *)
+                  networksetup -listallnetworkservices \\
+                    | grep -q "$name" \\
+                      || networksetup -createnetworkservice "$name" "$name" 2>/dev/null \\
+                      || continue
+                        ;;
+              esac
+            done
+
         networksetup -listnetworkservices | grep -q "Thunderbolt Bridge" && {
           networksetup -setnetworkserviceenabled "Thunderbolt Bridge" off
         } || true

dashboard/parts.nix

@@ -3,12 +3,28 @@
   perSystem =
     { pkgs, lib, ... }:
     let
+      # Filter source to ONLY include package.json and package-lock.json
+      # This ensures prettier-svelte only rebuilds when lockfiles change
+      dashboardLockfileSrc = lib.cleanSourceWith {
+        src = inputs.self;
+        filter =
+          path: type:
+          let
+            baseName = builtins.baseNameOf path;
+            isDashboardDir = baseName == "dashboard" && type == "directory";
+            isPackageFile =
+              (lib.hasInfix "/dashboard/" path || lib.hasSuffix "/dashboard" (builtins.dirOf path))
+              && (baseName == "package.json" || baseName == "package-lock.json");
+          in
+          isDashboardDir || isPackageFile;
+      };
+
       # Stub source with lockfiles and minimal files for build to succeed
       # This allows prettier-svelte to avoid rebuilding when dashboard source changes
       dashboardStubSrc = pkgs.runCommand "dashboard-stub-src" { } ''
         mkdir -p $out
-        cp ${inputs.self}/dashboard/package.json $out/
-        cp ${inputs.self}/dashboard/package-lock.json $out/
+        cp ${dashboardLockfileSrc}/dashboard/package.json $out/
+        cp ${dashboardLockfileSrc}/dashboard/package-lock.json $out/
         # Minimal files so vite build succeeds (produces empty output)
         echo '<!DOCTYPE html><html><head></head><body></body></html>' > $out/index.html
         mkdir -p $out/src

rust/cluster_membership/Cargo.toml

@@ -0,0 +1,23 @@
+[package]
+name = "cluster_membership"
+version.workspace = true
+edition.workspace = true
+publish = false
+
+[dependencies]
+# util
+anyhow.workspace = true
+log.workspace = true
+tracing-subscriber = { version = "0.3.19", features = ["default", "env-filter"] }
+
+# async
+tokio = { workspace = true, features = ["full"] }
+futures-timer = { workspace = true }
+futures-lite = "2.6.1"
+
+# networking
+libp2p = { workspace = true, features = ["full"] }
+async-trait = "0.1.89"
+
+[lints]
+workspace = true

rust/cluster_membership/examples/chatroom.rs

@@ -0,0 +1,30 @@
+use cluster_membership::Peer;
+use libp2p::identity::ed25519::SecretKey;
+use tokio::io::{self, AsyncBufReadExt};
+use tracing_subscriber::{EnvFilter, filter::LevelFilter};
+
+#[tokio::main]
+async fn main() {
+    let _ = tracing_subscriber::fmt()
+        .with_env_filter(EnvFilter::from_default_env().add_directive(LevelFilter::INFO.into()))
+        .try_init();
+
+    let (mut peer, send, mut recv) =
+        Peer::new(SecretKey::generate(), "hello".to_string()).expect("peer should always build");
+
+    let ch = peer.subscribe("chatroom".to_string());
+    let jh = tokio::spawn(async move { peer.run().await });
+
+    let mut stdin = io::BufReader::new(io::stdin()).lines();
+    loop {
+        tokio::select! {
+            Ok(Some(line)) = stdin.next_line() => {send.send((ch.clone(), line.into_bytes())).await.expect("example");}
+            Some(r) = recv.recv() => match r {
+                Ok((_, id, line)) => println!("{:?}:{:?}", id, String::from_utf8_lossy(&line)),
+                Err(e) => eprintln!("{e:?}"),
+            },
+            else => break
+        }
+    }
+    jh.await.expect("task failure");
+}

rust/cluster_membership/src/lib.rs

@@ -0,0 +1,227 @@
+use libp2p::{
+    Multiaddr, PeerId, Swarm, SwarmBuilder,
+    futures::StreamExt,
+    gossipsub::{self, PublishError, Sha256Topic, TopicHash},
+    identify,
+    identity::{Keypair, ed25519},
+    mdns,
+    swarm::{NetworkBehaviour, SwarmEvent, dial_opts::DialOpts},
+};
+use std::{
+    collections::HashMap,
+    time::{Duration, Instant},
+};
+use tokio::{select, sync::mpsc};
+
+const DEFAULT_BUFFER_SIZE: usize = 10;
+const MDNS_IGNORE_DURATION_SECS: u64 = 30;
+
+impl Peer {
+    pub fn new(
+        identity: ed25519::SecretKey,
+        namespace: String,
+    ) -> anyhow::Result<(
+        Self,
+        mpsc::Sender<(TopicHash, Vec<u8>)>,
+        mpsc::Receiver<Result<(TopicHash, PeerId, Vec<u8>), PublishError>>,
+    )> {
+        let mut id_bytes = identity.as_ref().to_vec();
+
+        let mut swarm =
+            SwarmBuilder::with_existing_identity(Keypair::ed25519_from_bytes(&mut id_bytes)?)
+                .with_tokio()
+                .with_quic()
+                // TODO(evan): .with_bandwidth_metrics();
+                .with_behaviour(|kp| Behaviour::new(kp, namespace.clone()))?
+                .build();
+
+        swarm.listen_on("/ip6/::/udp/0/quic-v1".parse()?)?;
+        swarm.listen_on("/ip4/0.0.0.0/udp/0/quic-v1".parse()?)?;
+        let (to_swarm, from_client) = mpsc::channel(DEFAULT_BUFFER_SIZE);
+        let (to_client, from_swarm) = mpsc::channel(DEFAULT_BUFFER_SIZE);
+        Ok((
+            Self {
+                swarm,
+                namespace,
+                denied: HashMap::new(),
+                from_client,
+                to_client,
+            },
+            to_swarm,
+            from_swarm,
+        ))
+    }
+
+    pub fn subscribe(&mut self, topic: String) -> TopicHash {
+        let topic = Sha256Topic::new(topic);
+        self.swarm
+            .behaviour_mut()
+            .gossipsub
+            .subscribe(&topic)
+            .expect("topic filtered");
+        topic.hash()
+    }
+
+    pub async fn run(&mut self) {
+        loop {
+            select! {
+                ev = self.swarm.select_next_some() => {
+                    let Ok(()) = self.handle_swarm_event(ev).await else {
+                        return
+                    };
+                },
+                Some(msg) = self.from_client.recv() => {
+                    if let Err(e) = self.swarm.behaviour_mut().gossipsub.publish(msg.0, msg.1) {
+                        let Ok(()) = self.to_client.send(Err(e)).await else {
+                            return
+                        };
+                    }
+                },
+            }
+        }
+    }
+
+    async fn handle_swarm_event(&mut self, event: SwarmEvent<BehaviourEvent>) -> Result<(), ()> {
+        let SwarmEvent::Behaviour(event) = event else {
+            if let SwarmEvent::NewListenAddr {
+                listener_id: _,
+                address,
+            } = event
+            {
+                log::info!("new listen address {address}")
+            }
+            return Ok(());
+        };
+        match event {
+            BehaviourEvent::Mdns(mdns_event) => match mdns_event {
+                mdns::Event::Discovered(vec) => {
+                    // Dial everyone
+                    let mut addrs = HashMap::<PeerId, Vec<Multiaddr>>::new();
+                    vec.into_iter()
+                        .filter(|(peer_id, _)| {
+                            self.denied.get(peer_id).is_none_or(|t| {
+                                t.elapsed() > Duration::from_secs(MDNS_IGNORE_DURATION_SECS)
+                            })
+                        })
+                        .for_each(|(peer_id, addr)| addrs.entry(peer_id).or_default().push(addr));
+                    addrs.into_iter().for_each(|(peer_id, addrs)| {
+                        let _ = self
+                            .swarm
+                            .dial(DialOpts::peer_id(peer_id).addresses(addrs).build());
+                    });
+                }
+                mdns::Event::Expired(vec) => {
+                    vec.iter().for_each(|(peer_id, _)| {
+                        log::debug!("{peer_id} no longer reachable on mDNS");
+                        self.swarm
+                            .behaviour_mut()
+                            .gossipsub
+                            .remove_explicit_peer(peer_id);
+                    });
+                }
+            },
+            BehaviourEvent::Identify(identify::Event::Received {
+                connection_id: _,
+                peer_id,
+                info,
+            }) => {
+                if info
+                    .protocols
+                    .iter()
+                    .any(|p| p.as_ref().contains(&self.namespace))
+                {
+                    self.passed_namespace(peer_id);
+                } else {
+                    self.failed_namespace(peer_id);
+                }
+            }
+            BehaviourEvent::Gossipsub(gossipsub::Event::Message {
+                propagation_source: _,
+                message_id: _,
+                message:
+                    gossipsub::Message {
+                        topic,
+                        data,
+                        source: Some(source_peer),
+                        ..
+                    },
+            }) => {
+                let Ok(()) = self.to_client.send(Ok((topic, source_peer, data))).await else {
+                    return Err(());
+                };
+            }
+            _ => {}
+        }
+        Ok(())
+    }
+
+    fn passed_namespace(&mut self, peer: PeerId) {
+        log::info!("new peer {peer:?}");
+        self.denied.remove(&peer);
+        self.swarm
+            .behaviour_mut()
+            .gossipsub
+            .remove_blacklisted_peer(&peer);
+        self.swarm
+            .behaviour_mut()
+            .gossipsub
+            .add_explicit_peer(&peer);
+    }
+
+    fn failed_namespace(&mut self, peer: PeerId) {
+        log::debug!("{peer} failed handshake");
+        self.denied.insert(peer, Instant::now());
+        self.swarm.behaviour_mut().gossipsub.blacklist_peer(&peer);
+        // we don't care if disconnect fails
+        let _ = self.swarm.disconnect_peer_id(peer);
+    }
+}
+
+pub struct Peer {
+    pub swarm: Swarm<Behaviour>,
+    denied: HashMap<PeerId, Instant>,
+    namespace: String,
+    to_client: mpsc::Sender<Result<(TopicHash, PeerId, Vec<u8>), PublishError>>,
+    from_client: mpsc::Receiver<(TopicHash, Vec<u8>)>,
+}
+
+#[test]
+fn foo() {
+    fn bar<T: Send>(t: T) {}
+    let p: Peer = unimplemented!();
+    bar(p);
+}
+
+#[derive(NetworkBehaviour)]
+pub struct Behaviour {
+    mdns: mdns::tokio::Behaviour,
+    pub gossipsub: gossipsub::Behaviour,
+    identify: identify::Behaviour,
+}
+
+impl Behaviour {
+    fn new(kp: &Keypair, namespace: String) -> Self {
+        let mdns = mdns::tokio::Behaviour::new(Default::default(), kp.public().to_peer_id())
+            .expect("implementation is infallible");
+        let gossipsub = gossipsub::Behaviour::new(
+            gossipsub::MessageAuthenticity::Signed(kp.clone()),
+            gossipsub::ConfigBuilder::default()
+                .max_transmit_size(1024 * 1024)
+                .protocol_id_prefix(format!("/exo/gossip/{namespace}/v1"))
+                .build()
+                .expect("fixed gossipsub config should always build"),
+        )
+        .expect("fixed gossipsub init should always build");
+
+        let identify = identify::Behaviour::new(
+            identify::Config::new_with_signed_peer_record(format!("/exo/identity/v1"), kp)
+                .with_push_listen_addr_updates(true),
+        );
+
+        Behaviour {
+            mdns,
+            gossipsub,
+            identify,
+        }
+    }
+}

rust/exo_pyo3_bindings/Cargo.toml

@@ -22,6 +22,7 @@ doc = false
 workspace = true
 
 [dependencies]
+cluster_membership.workspace = true
 networking = { workspace = true }
 
 # interop

rust/exo_pyo3_bindings/src/pylibp2p/mod.rs

@@ -6,3 +6,41 @@
 
 pub mod ident;
 pub mod multiaddr;
+
+use std::sync::Mutex;
+
+use cluster_membership::Peer;
+use libp2p::identity::ed25519::Keypair;
+use pyo3::prelude::*;
+use pyo3_stub_gen::derive::{gen_stub_pyclass, gen_stub_pymethods};
+
+#[gen_stub_pyclass]
+#[pyclass]
+#[derive(Clone)]
+pub struct PyKeypair(Keypair);
+
+#[gen_stub_pymethods]
+#[pymethods]
+impl PyKeypair {
+    #[staticmethod]
+    fn generate() -> Self {
+        Self(Keypair::generate())
+    }
+}
+
+#[gen_stub_pyclass]
+#[pyclass]
+pub struct PyPeer(Mutex<Peer>);
+
+#[gen_stub_pymethods]
+#[pymethods]
+impl PyPeer {
+    #[staticmethod]
+    fn init(kp: PyKeypair, namespace: String) -> PyResult<Self> {
+        Ok(PyPeer(Mutex::new(
+            Peer::new(kp.0.secret(), namespace)
+                .map_err(|e| e.pyerr())?
+                .0,
+        )))
+    }
+}

src/exo/utils/info_gatherer/info_gatherer.py

@@ -349,13 +349,8 @@ class InfoGatherer:
     async def _monitor_misc(self):
         if self.misc_poll_interval is None:
             return
-        prev = await MiscData.gather()
-        await self.info_sender.send(prev)
         while True:
-            curr = await MiscData.gather()
-            if prev != curr:
-                prev = curr
-                await self.info_sender.send(curr)
+            await self.info_sender.send(await MiscData.gather())
             await anyio.sleep(self.misc_poll_interval)
 
     async def _monitor_system_profiler_thunderbolt_data(self):
@@ -365,15 +360,12 @@ class InfoGatherer:
         if iface_map is None:
             return
 
-        old_idents = []
         while True:
             data = await ThunderboltConnectivity.gather()
             assert data is not None
 
             idents = [it for i in data if (it := i.ident(iface_map)) is not None]
-            if idents != old_idents:
-                await self.info_sender.send(MacThunderboltIdentifiers(idents=idents))
-            old_idents = idents
+            await self.info_sender.send(MacThunderboltIdentifiers(idents=idents))
 
             conns = [it for i in data if (it := i.conn()) is not None]
             await self.info_sender.send(MacThunderboltConnections(conns=conns))
@@ -398,22 +390,17 @@ class InfoGatherer:
     async def _watch_system_info(self):
         if self.interface_watcher_interval is None:
             return
-        old_nics = []
         while True:
             nics = await get_network_interfaces()
-            if nics != old_nics:
-                old_nics = nics
-                await self.info_sender.send(NodeNetworkInterfaces(ifaces=nics))
+            await self.info_sender.send(NodeNetworkInterfaces(ifaces=nics))
             await anyio.sleep(self.interface_watcher_interval)
 
     async def _monitor_thunderbolt_bridge_status(self):
         if self.thunderbolt_bridge_poll_interval is None:
             return
-        prev: ThunderboltBridgeInfo | None = None
         while True:
             curr = await ThunderboltBridgeInfo.gather()
-            if curr is not None and prev != curr:
-                prev = curr
+            if curr is not None:
                 await self.info_sender.send(curr)
             await anyio.sleep(self.thunderbolt_bridge_poll_interval)