A node is a connected computer or server running dedicated protocol software to communicate, validate transactions, and maintain ledger data across a decentralized blockchain network.
Core Functions and Network Roles
Nodes form the peer-to-peer infrastructure that keeps a blockchain decentralized. When a user submits a transfer or smart contract execution, the transaction broadcasts across the network from one peer to another. Each validating node independently checks the cryptographic signature, account balance, and transaction fee against protocol rules before relaying it to the local memory pool.
Different node variants fulfill distinct responsibilities across the architecture:
- Full nodes: Store complete rule sets and validate every block and transaction independently, serving as the foundational verification layer for the protocol.
- Archival nodes: Retain the entire historical record of all past states, enabling comprehensive historical balance lookups and deep data analysis.
- Light nodes: Download only block headers using Simplified Payment Verification mechanisms, relying on full nodes to verify transaction inclusion without storing large ledgers.
All miners and block proposers run node software to track state and broadcast proposed blocks. However, standard full nodes operate strictly to audit and propagate network activity without creating new blocks or earning protocol block subsidies directly.
Pruning, Outages, and Architectural Edge Cases
Operating or interacting with a node introduces practical edge cases that affect data availability and state verification:
- Pruned storage limitations: Pruned full nodes verify the entire history from genesis but discard older transaction data once validated to save disk space. A pruned node cannot serve historical blockchain state queries to other peers.
- Network isolation and desynchronization: If a node drops offline or suffers latency degradation, it temporarily falls out of sync with the canonical chain tip. Transactions submitted through an out-of-sync node risk relying on stale state.
- Remote Procedure Call provider dependency: Many decentralized applications rely on centralized third-party node providers rather than self-hosted instances. While reducing operational overhead, this introduces potential downtime and single-point-of-failure risks.
- Fork divergence: During consensus splits or soft forks, nodes running outdated software versions may misinterpret block validity, failing to follow the active chain consensus rules.