Peer Discovery

Overview

Peer discovery supplies a node with candidate peers before a Peer Connection can be established. The discovered record usually contains a network address and, in identity-based systems, a Peer ID. A new node may begin with hard-coded bootstrap peers or DNS seeds, then learn additional addresses from connected peers, a distributed hash table, routing records, or local multicast discovery.

Discovery is separate from dialing and trust. A discovered address may be offline, unreachable behind NAT, incompatible, or malicious. The node still needs to establish a secure connection, authenticate the peer where supported, negotiate protocols, and validate all received data. Discovery systems also need expiry and freshness rules because peer addresses change and stale records can waste connection attempts.

Security depends on obtaining candidates from diverse sources. If an attacker controls most discovery responses, it can steer a node toward an isolated Peer Set and attempt an eclipse attack. Implementations may limit how many addresses are accepted from one source, query disjoint paths, randomize selections, and preserve previously successful peers. Privacy-sensitive nodes may also restrict address advertisement or use anonymity networks.

Operators should monitor discovery success rate, unique candidate count, address diversity, dial outcomes, and time required for Node Bootstrapping. A node that repeatedly connects to the same small group may appear healthy while having poor network visibility. Reliable payment and indexing systems often combine protocol discovery with curated fallback peers and multiple independent nodes, without treating any discovery source as authoritative.

Audit and recovery workflows should preserve the discovery source, candidate identity and addresses, freshness, dial outcome, and diversity characteristics. These records help because an attacker controlling discovery can steer a node toward an isolated neighborhood without breaking transport security. Production safeguards need to query independent sources, expire stale records, and measure whether successful peers remain diverse.