dns bootstrap

[dzdidi]

DNS bootstrap instead of manual

Blocked by / Based on #6

[Copilot]

Pull request overview

Adds DNS-based peer discovery (BOLT-0010) so the node can bootstrap outbound connections automatically from well-known DNS seeds instead of relying solely on manual peer configuration.

Changes:

• Introduces a new dns_bootstrap module to query SRV records, decode bech32 pubkeys from hostnames, and resolve A/AAAA records into connectable peers.
• Wires optional dns_bootstrap configuration through config.json → NodeConfig → LdkUserInfo and starts a periodic bootstrap task in start_ldk.
• Adds hickory-resolver dependency to perform async DNS lookups.

Reviewed changes

Copilot reviewed 9 out of 10 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
src/main.rs	Starts periodic DNS bootstrap task to discover/connect peers.
src/dns_bootstrap/bootstrapper.rs	Core bootstrap logic: SRV query → pubkey decode → host resolve → peer list.
src/dns_bootstrap/query.rs	DNS resolver wrapper for SRV + IP lookups with timeouts.
src/dns_bootstrap/pubkey.rs	Decodes node pubkeys from bech32 DNS labels; includes unit tests.
src/dns_bootstrap/config.rs	Defines DnsBootstrapConfig + default seed list.
src/dns_bootstrap/mod.rs	Module exports and error type definitions.
src/config.rs	Adds dns_bootstrap to NodeConfig and config help output.
src/cli.rs	Extends LdkUserInfo to carry dns_bootstrap config.
Cargo.toml	Adds hickory-resolver dependency.
Cargo.lock	Lockfile updates for new DNS dependency graph.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

query_name unconditionally appends a trailing dot, so a config value like nodes.lightning.wiki. becomes _nodes._tcp.nodes.lightning.wiki.. and may fail to resolve. Normalize seed (e.g., trim trailing '.') before formatting, and/or validate seed strings during config parsing.

Suggested change

	let query_name = format!("_nodes._tcp.{}.", seed);
	let normalized_seed = seed.trim_end_matches('.');
	let query_name = format!("_nodes._tcp.{}.", normalized_seed);

[Copilot]

sample_node_addrs may include duplicates (same node returned multiple times across SRV records or across seeds), leading to redundant connection attempts. Consider tracking a seen set (e.g., HashSet<NodeId> or HashSet<(NodeId, SocketAddr)>) and skipping duplicates before pushing into peers.

[Copilot]

New DNS bootstrap behavior isn’t covered by tests (selection/ignore/dedup/error handling). Consider making the DNS query layer injectable (e.g., via a trait) so DnsBootstrapper::sample_node_addrs can be unit-tested deterministically without real DNS.

[Copilot]

interval_secs comes from config and can be set to 0, but tokio::time::interval(Duration::from_secs(0)) panics. Add validation (e.g., reject 0 with a clear error) before spawning the task, or clamp to a minimum of 1s so misconfiguration can’t crash the bootstrap task.

Suggested change

	println!(
	"DNS bootstrap enabled: {} seeds, target {} peers, interval {}s",
	bootstrapper.config().seeds.len(),
	num_peers,
	interval_secs,
	);
	tokio::spawn(async move {
	// Initial delay to let networking and gossip start.
	tokio::time::sleep(Duration::from_secs(5)).await;
	let mut interval =
	tokio::time::interval(Duration::from_secs(interval_secs));
	loop {
	interval.tick().await;
	if stop_bootstrap.load(Ordering::Acquire) {
	return;
	}
	// Build ignore set from currently connected peers.
	let ignore: std::collections::HashSet<
	lightning::routing::gossip::NodeId,
	> = bootstrap_pm
	.list_peers()
	.iter()
	.map(|p| {
	lightning::routing::gossip::NodeId::from_pubkey(
	&p.counterparty_node_id,
	)
	})
	.collect();
	match bootstrapper.sample_node_addrs(num_peers, &ignore).await {
	Ok(peers) => {
	println!(
	"[dns_bootstrap] Discovered {} peers",
	peers.len()
	);
	for peer in peers {
	let _ = cli::do_connect_peer(
	peer.pubkey,
	peer.addr,
	Arc::clone(&bootstrap_pm),
	)
	.await;
	}
	},
	Err(e) => {
	eprintln!("[dns_bootstrap] Bootstrap failed: {}", e);
	},
	}
	}
	});
	if interval_secs == 0 {
	eprintln!(
	"[dns_bootstrap] Invalid configuration: interval_secs must be greater than 0; DNS bootstrap disabled."
	);
	} else {
	println!(
	"DNS bootstrap enabled: {} seeds, target {} peers, interval {}s",
	bootstrapper.config().seeds.len(),
	num_peers,
	interval_secs,
	);
	tokio::spawn(async move {
	// Initial delay to let networking and gossip start.
	tokio::time::sleep(Duration::from_secs(5)).await;
	let mut interval =
	tokio::time::interval(Duration::from_secs(interval_secs));
	loop {
	interval.tick().await;
	if stop_bootstrap.load(Ordering::Acquire) {
	return;
	}
	// Build ignore set from currently connected peers.
	let ignore: std::collections::HashSet<
	lightning::routing::gossip::NodeId,
	> = bootstrap_pm
	.list_peers()
	.iter()
	.map(|p| {
	lightning::routing::gossip::NodeId::from_pubkey(
	&p.counterparty_node_id,
	)
	})
	.collect();
	match bootstrapper.sample_node_addrs(num_peers, &ignore).await {
	Ok(peers) => {
	println!(
	"[dns_bootstrap] Discovered {} peers",
	peers.len()
	);
	for peer in peers {
	let _ = cli::do_connect_peer(
	peer.pubkey,
	peer.addr,
	Arc::clone(&bootstrap_pm),
	)
	.await;
	}
	},
	Err(e) => {
	eprintln!("[dns_bootstrap] Bootstrap failed: {}", e);
	},
	}
	}
	});
	}

[Copilot]

fallback_soa is defined and documented as a TCP fallback mechanism but is never used anywhere in the bootstrapper/query logic. Either implement the SOA/TCP fallback behavior or remove this field (and the config help example) to avoid exposing non-functional configuration.