archy/core/archipelago/src/server.rs

use crate::api::ApiHandler;
use crate::config::{Config, ContainerRuntime};
use crate::container::{docker_packages, DockerPackageScanner};
use crate::identity::{self, NodeIdentity};
use crate::monitoring::MetricsStore;
use crate::node_message;
use crate::nostr_discovery;
use crate::nostr_handshake;
use crate::peers;
use crate::state::StateManager;
use anyhow::Result;
use hyper::server::conn::Http;
use hyper::service::service_fn;
use std::collections::HashMap;
use std::net::SocketAddr;
use std::sync::Arc;
use std::time::Duration;
use tokio::net::TcpListener;
use tracing::{debug, error, info, warn};

pub struct Server {
    _config: Config,
    _identity: Arc<NodeIdentity>,
    api_handler: Arc<ApiHandler>,
    _state_manager: Arc<StateManager>,
}

impl Server {
    pub async fn new(config: Config) -> Result<Self> {
        let state_manager = Arc::new(StateManager::new());

        // Load node identity and set stable server_info.
        // Detect seed-backed vs legacy vs fresh install.
        let identity_dir = config.data_dir.join("identity");
        let has_seed = crate::seed::seed_exists(&config.data_dir);
        let has_node_key = NodeIdentity::key_exists(&identity_dir);

        let identity = if has_node_key {
            // Existing keys on disk (seed-derived or legacy random) — load them.
            NodeIdentity::load_or_create(&identity_dir).await?
        } else {
            // Fresh install — create a temporary identity.
            // Onboarding will overwrite this with seed-derived keys.
            NodeIdentity::load_or_create(&identity_dir).await?
        };

        let (mut data, _) = state_manager.get_snapshot().await;
        data.server_info.id = identity.node_id();
        data.server_info.pubkey = identity.pubkey_hex();
        data.server_info.seed_backed = has_seed;
        // Load persisted server name
        let name_file = config.data_dir.join("server-name");
        if let Ok(name) = tokio::fs::read_to_string(&name_file).await {
            let name = name.trim().to_string();
            if !name.is_empty() {
                data.server_info.name = Some(name);
            }
        }
        data.server_info.tor_address = docker_packages::read_tor_address("archipelago").await;
        if let Some(ref tor) = data.server_info.tor_address {
            data.server_info.node_address = Some(identity.node_address(tor));
        }
        state_manager.update_data(data.clone()).await;

        // Retry Tor address in background — Tor may not be ready at startup
        if data.server_info.tor_address.is_none() {
            let sm = state_manager.clone();
            let pubkey = identity.pubkey_hex();
            tokio::spawn(async move {
                for delay in [5, 10, 20, 30, 60] {
                    tokio::time::sleep(std::time::Duration::from_secs(delay)).await;
                    if let Some(tor) = docker_packages::read_tor_address("archipelago").await {
                        let (mut d, _) = sm.get_snapshot().await;
                        let addr =
                            format!("archipelago://{}#{}", tor.trim_end_matches('/'), pubkey);
                        d.server_info.tor_address = Some(tor.clone());
                        d.server_info.node_address = Some(addr);
                        sm.update_data(d).await;
                        tracing::info!(
                            "Tor address discovered after startup: {}",
                            &tor[..20.min(tor.len())]
                        );
                        break;
                    }
                }
            });
        }

        // Load persisted messages (Archipelago channel)
        node_message::init(&config.data_dir).await;

        // Auto-create default identity if none exist (fresh boot or factory reset)
        {
            let im = crate::identity_manager::IdentityManager::new(&config.data_dir).await;
            if let Ok(mgr) = im {
                if let Ok((list, _)) = mgr.list().await {
                    if list.is_empty() {
                        match mgr
                            .create(
                                "Default".to_string(),
                                crate::identity_manager::IdentityPurpose::Personal,
                            )
                            .await
                        {
                            Ok(record) => {
                                let _ = mgr.create_nostr_key(&record.id).await;
                                tracing::info!(did = %record.did, "Auto-created default identity with Nostr key");
                            }
                            Err(e) => tracing::debug!("Auto-identity creation (non-fatal): {}", e),
                        }
                    }
                }
            }
        }

        // Revoke any previously published Nostr data (runs before publish so revocation is not overwritten)
        let identity_dir = config.data_dir.join("identity");
        let tor_proxy_revoke = config.nostr_tor_proxy.clone();
        if let Err(e) =
            nostr_discovery::revoke_if_needed(&identity_dir, tor_proxy_revoke.as_deref()).await
        {
            tracing::debug!("Nostr revoke (non-fatal): {}", e);
        }

        // Publish presence-only to Nostr (DID + Nostr pubkey, NO onion address).
        // Onion addresses are exchanged privately via NIP-44 encrypted DMs.
        if config.nostr_discovery_enabled && !config.nostr_relays.is_empty() {
            let identity_dir = config.data_dir.join("identity");
            let did =
                identity::did_key_from_pubkey_hex(&data.server_info.pubkey).unwrap_or_default();
            let version = data.server_info.version.clone();
            let relays = config.nostr_relays.clone();
            let tor_proxy = config.nostr_tor_proxy.clone();
            tokio::spawn(async move {
                if let Err(e) = nostr_handshake::publish_presence(
                    &identity_dir,
                    &did,
                    &version,
                    &relays,
                    tor_proxy.as_deref(),
                )
                .await
                {
                    tracing::debug!("Nostr presence publish (non-fatal): {}", e);
                }
            });
        }
        info!(
            "🔑 Node identity: {} (pubkey: {}...)",
            identity.node_id(),
            &identity.pubkey_hex()[..16.min(identity.pubkey_hex().len())]
        );

        let identity = Arc::new(identity);

        // Create metrics store and spawn background collector
        let metrics_store = Arc::new(MetricsStore::with_data_dir(config.data_dir.clone()).await);
        let metrics_for_telemetry = metrics_store.clone();
        crate::monitoring::spawn_metrics_collector(
            metrics_store.clone(),
            Some(state_manager.clone()),
            Some(config.data_dir.clone()),
        );

        let api_handler =
            Arc::new(ApiHandler::new(config.clone(), state_manager.clone(), metrics_store).await?);

        // Initialize mesh networking service (if config has enabled: true)
        {
            let data_dir = config.data_dir.clone();
            let did =
                identity::did_key_from_pubkey_hex(&data.server_info.pubkey).unwrap_or_default();
            let pubkey_hex = identity.pubkey_hex();
            let signing_key = identity.signing_key();
            match crate::mesh::MeshService::new(&data_dir, signing_key, &did, &pubkey_hex).await {
                Ok(mut mesh_service) => {
                    // Pass the human-readable server name for mesh adverts
                    mesh_service.set_server_name(data.server_info.name.clone());
                    let mut mesh_config = crate::mesh::load_config(&data_dir)
                        .await
                        .unwrap_or_default();

                    // Auto-enable mesh if a radio is detected and no config exists yet
                    if !mesh_config.enabled {
                        let devices = crate::mesh::detect_devices().await;
                        if !devices.is_empty() {
                            info!("📡 Auto-detected mesh radio: {:?} — enabling mesh", devices);
                            mesh_config.enabled = true;
                            mesh_config.device_path = Some(devices[0].clone());
                            let _ = crate::mesh::save_config(&data_dir, &mesh_config).await;
                        }
                    }

                    if mesh_config.enabled {
                        if let Err(e) = mesh_service.start() {
                            warn!("Mesh service start failed (non-fatal): {}", e);
                        } else {
                            info!("📡 Mesh networking started");
                        }
                    }
                    api_handler
                        .rpc_handler()
                        .set_mesh_service(mesh_service)
                        .await;
                    info!("📡 Mesh service initialized");
                }
                Err(e) => {
                    warn!("Mesh service init failed (non-fatal): {}", e);
                }
            }
        }

        // Initialize transport router (unified routing: mesh > lan > tor)
        {
            let data_dir = config.data_dir.clone();
            let did =
                identity::did_key_from_pubkey_hex(&data.server_info.pubkey).unwrap_or_default();
            let pubkey_hex = identity.pubkey_hex();
            let mesh_config = crate::mesh::load_config(&data_dir)
                .await
                .unwrap_or_default();
            let mesh_only = mesh_config.mesh_only_mode.unwrap_or(false);

            match crate::transport::PeerRegistry::load(&data_dir).await {
                Ok(registry) => {
                    let registry = std::sync::Arc::new(registry);
                    let mut transports: Vec<Box<dyn crate::transport::NodeTransport>> = Vec::new();

                    // Tor transport (always register — availability checked dynamically)
                    transports.push(Box::new(crate::transport::tor::TorTransport::new(
                        &pubkey_hex,
                    )));

                    // Mesh transport (wraps the mesh service)
                    transports.push(Box::new(
                        crate::transport::mesh_transport::MeshTransport::new(
                            api_handler.rpc_handler().mesh_service_arc(),
                        ),
                    ));

                    // LAN transport (mDNS discovery)
                    let mut lan = crate::transport::lan::LanTransport::new(&did, &pubkey_hex, 5678);
                    match lan.start(registry.clone()) {
                        Ok(()) => info!("📡 LAN transport (mDNS) started"),
                        Err(e) => debug!("LAN transport init (non-fatal): {}", e),
                    }
                    transports.push(Box::new(lan));

                    let router = std::sync::Arc::new(crate::transport::TransportRouter::new(
                        transports, registry, mesh_only,
                    ));
                    api_handler.rpc_handler().set_transport_router(router).await;
                    info!("📡 Transport router initialized (mesh_only={})", mesh_only);
                }
                Err(e) => {
                    warn!("Transport router init failed (non-fatal): {}", e);
                }
            }
        }

        // Register Archipelago DWN protocols (background, non-blocking)
        {
            let data_dir = config.data_dir.clone();
            tokio::spawn(async move {
                if let Err(e) = register_dwn_protocols(&data_dir).await {
                    debug!("DWN protocol registration (non-fatal): {}", e);
                }
            });
        }

        // Periodic Tor address refresh (runs regardless of dev_mode)
        // Picks up hostname when Tor creates it after startup/rotation (30-60s delay)
        {
            let state = state_manager.clone();
            let identity_clone = identity.clone();
            tokio::spawn(async move {
                let mut interval = tokio::time::interval(Duration::from_secs(30));
                loop {
                    interval.tick().await;
                    if let Err(e) = refresh_tor_address(&state, identity_clone.as_ref()).await {
                        debug!("Tor address refresh (non-fatal): {}", e);
                    }
                }
            });
        }

        // Initialize container scanner — discovers installed apps from Podman/Docker
        {
            let scanner = create_docker_scanner(&config).await?;
            let state = state_manager.clone();
            let identity_clone = identity.clone();

            // Initial scan (delayed to let crash recovery finish first)
            tokio::spawn(async move {
                // Brief delay for containers to stabilize after boot
                tokio::time::sleep(Duration::from_secs(3)).await;
                info!("🐳 Scanning containers...");
                // Tracks how many consecutive scans each container has been absent from.
                // Prevents UI flapping when podman intermittently returns incomplete results.
                let mut absence_tracker: HashMap<String, u32> = HashMap::new();
                if let Err(e) = scan_and_update_packages(
                    &scanner,
                    &state,
                    identity_clone.as_ref(),
                    &mut absence_tracker,
                )
                .await
                {
                    error!("Failed to scan containers: {}", e);
                }

                // Periodic scan every 60 seconds (only broadcasts if state changed)
                // Uses an in-flight guard to skip scans when a previous one is still running
                let mut interval = tokio::time::interval(Duration::from_secs(60));
                // Skip missed ticks instead of catching up — prevents burst of scans
                // after a slow podman response (which causes DB lock storms)
                interval.set_missed_tick_behavior(tokio::time::MissedTickBehavior::Skip);
                let scanning = std::sync::Arc::new(std::sync::atomic::AtomicBool::new(false));
                loop {
                    interval.tick().await;
                    if scanning.load(std::sync::atomic::Ordering::Relaxed) {
                        debug!("Skipping container scan — previous scan still in progress");
                        continue;
                    }
                    scanning.store(true, std::sync::atomic::Ordering::Relaxed);
                    if let Err(e) = scan_and_update_packages(
                        &scanner,
                        &state,
                        identity_clone.as_ref(),
                        &mut absence_tracker,
                    )
                    .await
                    {
                        error!("Failed to update containers: {}", e);
                    }
                    scanning.store(false, std::sync::atomic::Ordering::Relaxed);
                }
            });
        }

        // Peer health monitoring — check every 5 minutes
        {
            let state = state_manager.clone();
            let data_dir = config.data_dir.clone();
            tokio::spawn(async move {
                let mut interval = tokio::time::interval(Duration::from_secs(300));
                loop {
                    interval.tick().await;
                    if let Err(e) = check_peer_health(&state, &data_dir).await {
                        debug!("Peer health check (non-fatal): {}", e);
                    }
                }
            });
        }

        // did:dht auto-refresh — re-publish DHT records every 2 hours
        if config.nostr_discovery_enabled {
            let data_dir = config.data_dir.clone();
            tokio::spawn(async move {
                let mut interval = tokio::time::interval(Duration::from_secs(7200));
                loop {
                    interval.tick().await;
                    let identity_dir = data_dir.join("identity");
                    let node_key_path = identity_dir.join("node_key");
                    if !node_key_path.exists() {
                        continue;
                    }
                    match tokio::fs::read(&node_key_path).await {
                        Ok(key_bytes) if key_bytes.len() == 32 => {
                            let mut seed = [0u8; 32];
                            seed.copy_from_slice(&key_bytes);
                            let signing_key = ed25519_dalek::SigningKey::from_bytes(&seed);
                            match crate::network::did_dht::create_and_publish(&signing_key, &[])
                                .await
                            {
                                Ok(did) => tracing::info!(did = %did, "did:dht record refreshed"),
                                Err(e) => tracing::debug!("did:dht refresh (non-fatal): {}", e),
                            }
                        }
                        _ => {
                            tracing::debug!("did:dht refresh skipped: no valid node key");
                        }
                    }
                }
            });
        }

        // Periodic federation state sync — every 30 min we call
        // federation::sync_with_peer on each Trusted peer. Without this
        // users had to manually click Sync for `fips_npub`/transport
        // badge/state updates to propagate; now it happens in the
        // background. Staggers peers with a 5s delay so we don't thunder
        // the Tor SOCKS proxy. Sync itself already prefers FIPS.
        {
            let data_dir = config.data_dir.clone();
            let state = state_manager.clone();
            tokio::spawn(async move {
                // First run 60s after boot to let onboarding settle.
                tokio::time::sleep(Duration::from_secs(60)).await;
                let mut interval = tokio::time::interval(Duration::from_secs(1800));
                interval.set_missed_tick_behavior(tokio::time::MissedTickBehavior::Delay);
                loop {
                    interval.tick().await;
                    let Ok(nodes) = crate::federation::load_nodes(&data_dir).await else {
                        continue;
                    };
                    if nodes.is_empty() {
                        continue;
                    }
                    let (data, _) = state.get_snapshot().await;
                    let Ok(local_did) =
                        crate::identity::did_key_from_pubkey_hex(&data.server_info.pubkey)
                    else {
                        continue;
                    };
                    let identity_dir = data_dir.join("identity");
                    let Ok(node_identity) =
                        crate::identity::NodeIdentity::load_or_create(&identity_dir).await
                    else {
                        continue;
                    };

                    for node in &nodes {
                        if node.trust_level == crate::federation::TrustLevel::Untrusted {
                            continue;
                        }
                        match crate::federation::sync_with_peer(
                            &data_dir,
                            node,
                            &local_did,
                            |bytes| node_identity.sign(bytes),
                        )
                        .await
                        {
                            Ok(_) => debug!(
                                "Periodic federation sync ok: {}",
                                node.did.chars().take(20).collect::<String>()
                            ),
                            Err(e) => debug!(
                                "Periodic federation sync with {}: {}",
                                node.did.chars().take(20).collect::<String>(),
                                e
                            ),
                        }
                        tokio::time::sleep(Duration::from_secs(5)).await;
                    }
                }
            });
        }

        // Container health monitoring — auto-restart unhealthy containers
        // Respects webhook config: skips when disabled or ContainerCrash not subscribed
        crate::health_monitor::spawn_health_monitor(state_manager.clone(), config.data_dir.clone());

        // Periodic telemetry reporter (every 15 min when opted in)
        crate::monitoring::spawn_telemetry_reporter(
            metrics_for_telemetry,
            Some(state_manager.clone()),
            config.data_dir.clone(),
        );

        Ok(Self {
            _config: config,
            _identity: identity,
            api_handler,
            _state_manager: state_manager,
        })
    }

    /// Serve with a graceful shutdown signal.
    ///
    /// `main_addr` is the primary listener (historically `127.0.0.1:5678`).
    /// The main listener always comes up on `main_addr`. The FIPS peer
    /// listener (path-filtered, bound to `fips0`'s ULA) is managed by a
    /// late-binding task that polls every 30s: if fips0 isn't up at
    /// startup (pre-onboarding install, legacy node pre-fips.install),
    /// it keeps trying until the interface appears — no archipelago
    /// restart required after the user activates FIPS.
    ///
    /// When `shutdown` completes, both listeners stop accepting and drain
    /// in-flight requests (bounded by `DRAIN_TIMEOUT`).
    pub async fn serve_with_shutdown(
        &self,
        main_addr: SocketAddr,
        shutdown: impl std::future::Future<Output = ()>,
    ) -> Result<()> {
        let active_connections = Arc::new(tokio::sync::Semaphore::new(1024));
        let (tx, rx_main) = tokio::sync::watch::channel(false);

        let main_task = tokio::spawn(accept_loop(
            self.api_handler.clone(),
            TcpListener::bind(main_addr).await?,
            active_connections.clone(),
            false, // main listener: no path filter
            rx_main,
            main_addr,
        ));

        // Peer listener: late-binding so we don't need an archipelago
        // restart when fips0 comes up after onboarding.
        let peer_task = tokio::spawn(peer_late_bind_loop(
            self.api_handler.clone(),
            active_connections.clone(),
            tx.subscribe(),
        ));

        shutdown.await;
        info!("Shutdown signal received, draining connections...");
        let _ = tx.send(true);

        // Wait up to 5s for in-flight requests.
        let drain_start = std::time::Instant::now();
        let drain_timeout = std::time::Duration::from_secs(5);
        while active_connections.available_permits() < 1024 {
            if drain_start.elapsed() > drain_timeout {
                warn!("Drain timeout reached, forcing shutdown");
                break;
            }
            tokio::time::sleep(std::time::Duration::from_millis(100)).await;
        }

        let _ = main_task.await;
        let _ = peer_task.await;

        info!("Shutdown complete");
        Ok(())
    }
}

/// Poll every 30s for `fips0`'s ULA; when it appears, bind the peer
/// listener and run the normal accept loop. If the bind fails (port
/// already taken, permissions), log and keep retrying. Returns on
/// shutdown. First tick fires immediately so the hot path for
/// already-up fips0 is still zero-cost.
async fn peer_late_bind_loop(
    handler: Arc<ApiHandler>,
    active_connections: Arc<tokio::sync::Semaphore>,
    mut shutdown_rx: tokio::sync::watch::Receiver<bool>,
) {
    let mut interval = tokio::time::interval(std::time::Duration::from_secs(30));
    interval.set_missed_tick_behavior(tokio::time::MissedTickBehavior::Delay);
    loop {
        tokio::select! {
            _ = interval.tick() => {
                let Some(ip) = crate::fips::iface::fips0_ula() else { continue };
                let addr = SocketAddr::new(
                    std::net::IpAddr::V6(ip),
                    crate::fips::dial::PEER_PORT,
                );
                let listener = match TcpListener::bind(addr).await {
                    Ok(l) => l,
                    Err(e) => {
                        warn!("FIPS peer listener bind {} failed: {} — retrying in 30s", addr, e);
                        continue;
                    }
                };
                info!("FIPS peer listener bound {}", addr);
                // Once bound, serve until shutdown fires. accept_loop
                // returns on shutdown, which also ends this outer loop.
                accept_loop(
                    handler,
                    listener,
                    active_connections,
                    true, // peer listener: apply path filter
                    shutdown_rx,
                    addr,
                )
                .await;
                return;
            }
            _ = shutdown_rx.changed() => {
                if *shutdown_rx.borrow() { return; }
            }
        }
    }
}

/// Whitelist of HTTP paths reachable via the peer-facing (FIPS) listener.
/// Every entry is an endpoint already protected by cryptographic auth
/// (ed25519 signature verification inside the handler, federation DID
/// headers checked by the content server, or JSON-RPC methods whose
/// handlers verify per-message signatures).
///
/// Anything not on this list returns 404 on the peer listener.
pub fn is_peer_allowed_path(path: &str) -> bool {
    // Exact matches
    matches!(
        path,
        "/health"
            | "/rpc/v1"
            | "/archipelago/node-message"
            | "/archipelago/mesh-typed"
            | "/dwn"
            | "/transport/inbox"
    )
    // Prefix-matched content endpoints (peer file browse + fetch)
        || path.starts_with("/content/")
}

async fn accept_loop(
    handler: Arc<ApiHandler>,
    listener: TcpListener,
    active_connections: Arc<tokio::sync::Semaphore>,
    peer_only: bool,
    mut shutdown_rx: tokio::sync::watch::Receiver<bool>,
    local_addr: SocketAddr,
) {
    loop {
        tokio::select! {
            result = listener.accept() => {
                let (stream, peer_addr) = match result {
                    Ok(c) => c,
                    Err(e) => {
                        error!("{} accept error: {}", local_addr, e);
                        continue;
                    }
                };
                let handler = handler.clone();
                let permit = active_connections.clone().acquire_owned().await;
                tokio::spawn(async move {
                    let _permit = permit;
                    let service = service_fn(move |req: hyper::Request<hyper::Body>| {
                        let handler = handler.clone();
                        async move {
                            if peer_only && !is_peer_allowed_path(req.uri().path()) {
                                let resp = hyper::Response::builder()
                                    .status(hyper::StatusCode::NOT_FOUND)
                                    .body(hyper::Body::empty())
                                    .expect("static response builds");
                                return Ok::<_, std::io::Error>(resp);
                            }
                            handler
                                .handle_request(req)
                                .await
                                .map_err(|e| std::io::Error::other(format!("{}", e)))
                        }
                    });
                    if let Err(e) = Http::new()
                        .http1_keep_alive(false)
                        .serve_connection(stream, service)
                        .with_upgrades()
                        .await
                    {
                        error!("Error serving connection from {}: {}", peer_addr, e);
                    }
                });
            }
            _ = shutdown_rx.changed() => {
                if *shutdown_rx.borrow() {
                    return;
                }
            }
        }
    }
}

async fn create_docker_scanner(config: &Config) -> Result<DockerPackageScanner> {
    let user = std::env::var("USER").unwrap_or_else(|_| "archipelago".to_string());

    let runtime: Arc<dyn archipelago_container::ContainerRuntime> = match &config.container_runtime
    {
        ContainerRuntime::Podman => {
            Arc::new(archipelago_container::PodmanRuntime::new(user.clone()))
        }
        ContainerRuntime::Docker => {
            Arc::new(archipelago_container::DockerRuntime::new(user.clone()))
        }
        ContainerRuntime::Auto => {
            Arc::new(archipelago_container::AutoRuntime::new(user.clone()).await?)
        }
    };

    Ok(DockerPackageScanner::new(runtime))
}

async fn refresh_tor_address(state: &StateManager, identity: &NodeIdentity) -> Result<()> {
    let tor_addr = docker_packages::read_tor_address("archipelago").await;
    let (current_data, _) = state.get_snapshot().await;
    if tor_addr != current_data.server_info.tor_address {
        let mut data = current_data;
        data.server_info.tor_address = tor_addr.clone();
        data.server_info.node_address = tor_addr.as_ref().map(|t| identity.node_address(t));
        state.update_data(data).await;
        if let Some(ref addr) = tor_addr {
            info!("🔒 Tor address updated: {}", addr);
        }
    }
    Ok(())
}

/// Number of consecutive absent scans before removing a container from state.
/// 3 scans × 30s = 90 seconds of absence before removal.
const CONTAINER_ABSENCE_THRESHOLD: u32 = 3;

async fn scan_and_update_packages(
    scanner: &DockerPackageScanner,
    state: &StateManager,
    identity: &NodeIdentity,
    absence_tracker: &mut HashMap<String, u32>,
) -> Result<()> {
    let packages = scanner.scan_containers().await?;

    let (current_data, _) = state.get_snapshot().await;
    let tor_addr = docker_packages::read_tor_address("archipelago").await;
    let tor_changed = tor_addr != current_data.server_info.tor_address;
    let first_scan = !current_data.server_info.status_info.containers_scanned;

    // Check if update scheduler has found an available update
    let update_available = crate::update::load_state(std::path::Path::new("/var/lib/archipelago"))
        .await
        .map(|s| s.available_update.is_some())
        .unwrap_or(false);
    let update_changed = update_available != current_data.server_info.status_info.updated;

    // Empty scan result = podman failure or timeout, preserve existing state
    if packages.is_empty() && !first_scan {
        if tor_changed || update_changed {
            let mut data = current_data;
            data.server_info.tor_address = tor_addr.clone();
            data.server_info.node_address = tor_addr.as_ref().map(|t| identity.node_address(t));
            data.server_info.status_info.updated = update_available;
            state.update_data(data).await;
        }
        return Ok(());
    }

    // Merge scan results with current state instead of full replacement.
    // This prevents containers from vanishing when podman intermittently
    // returns incomplete results under heavy load.
    let mut merged = current_data.package_data.clone();
    let mut changed = false;

    // Update/add containers found in this scan
    for (id, pkg) in &packages {
        absence_tracker.remove(id);
        if merged.get(id) != Some(pkg) {
            merged.insert(id.clone(), pkg.clone());
            changed = true;
        }
    }

    // Track containers in state but missing from this scan.
    // Only remove after CONTAINER_ABSENCE_THRESHOLD consecutive absent scans.
    let current_ids: Vec<String> = merged.keys().cloned().collect();
    for id in current_ids {
        if !packages.contains_key(&id) {
            let count = absence_tracker.entry(id.clone()).or_insert(0);
            *count += 1;
            if *count >= CONTAINER_ABSENCE_THRESHOLD {
                debug!(
                    "Removing {} from state after {} consecutive absent scans",
                    id, count
                );
                merged.remove(&id);
                absence_tracker.remove(&id);
                changed = true;
            }
        }
    }

    if changed || tor_changed || first_scan || update_changed {
        let mut data = current_data;
        data.package_data = merged;
        data.server_info.tor_address = tor_addr.clone();
        data.server_info.node_address = tor_addr.as_ref().map(|t| identity.node_address(t));
        data.server_info.status_info.containers_scanned = true;
        data.server_info.status_info.updated = update_available;
        state.update_data(data).await;
        debug!(
            "📦 State changed (packages={}, tor={}, first_scan={}, update={}), broadcasting update",
            changed, tor_changed, first_scan, update_changed
        );
    }

    Ok(())
}

/// Register Archipelago DWN protocols on startup.
async fn register_dwn_protocols(data_dir: &std::path::Path) -> Result<()> {
    use crate::network::dwn_store::{DwnStore, ProtocolDefinition};

    let protocols = [
        ("https://archipelago.dev/protocols/node-identity/v1", true),
        ("https://archipelago.dev/protocols/file-catalog/v1", true),
        ("https://archipelago.dev/protocols/federation/v1", false),
        ("https://archipelago.dev/protocols/app-deploy/v1", false),
    ];

    let store = DwnStore::new(data_dir).await?;
    let existing = store.list_protocols().await?;
    let existing_uris: std::collections::HashSet<String> =
        existing.iter().map(|p| p.protocol.clone()).collect();

    let mut registered = 0;
    for (uri, published) in &protocols {
        if existing_uris.contains(*uri) {
            continue;
        }
        let def = ProtocolDefinition {
            protocol: uri.to_string(),
            published: *published,
            types: std::collections::HashMap::new(),
            structure: std::collections::HashMap::new(),
            date_registered: chrono::Utc::now().to_rfc3339(),
        };
        store.register_protocol(&def).await?;
        registered += 1;
    }

    if registered > 0 {
        info!("📋 Registered {registered} DWN protocols");
    }
    Ok(())
}

/// Periodically check peer reachability and broadcast status changes.
async fn check_peer_health(state: &StateManager, data_dir: &std::path::Path) -> Result<()> {
    let known_peers = peers::load_peers(data_dir).await.unwrap_or_default();
    if known_peers.is_empty() {
        return Ok(());
    }

    let mut new_health = std::collections::HashMap::new();
    for peer in &known_peers {
        let fips_npub = crate::federation::fips_npub_for_onion(data_dir, &peer.onion).await;
        let reachable =
            node_message::check_peer_reachable(&peer.onion, fips_npub.as_deref())
                .await
                .unwrap_or(false);
        new_health.insert(peer.onion.clone(), reachable);
    }

    let (current_data, _) = state.get_snapshot().await;
    if current_data.peer_health != new_health {
        let mut data = current_data;
        data.peer_health = new_health;
        state.update_data(data).await;
        debug!("🔗 Peer health updated, broadcasting changes");
    }

    Ok(())
}