Node Availability

Definition

Node availability is the degree to which a blockchain node is reachable, synchronized, responsive, and capable of providing its intended service when required. It is broader than process uptime because a running node may be disconnected, stale, overloaded, or returning incorrect data.

Overview

Meaning and scope

Mechanism and operation

Node availability describes whether a node can perform its assigned function at a given time. A node process may be running while its database is corrupted, its peers are disconnected, or its chain head is far behind. An RPC node must answer queries accurately and within latency targets. A validator must communicate with peers and sign or propose when scheduled. An archival node must serve historical state, while a relay node must propagate transactions and blocks. One node can be available for basic health checks but unavailable for the particular operation a payment system needs.

Practical payment example

A blockchain provider returns successful health probes but fails most historical balance queries. Its SLA assessment therefore includes read capability and response accuracy, not only process uptime.

Operational checks

Useful signals include peer count, latest block age, finalized-block lag, RPC error rate, response latency, disk capacity, database status, and missed consensus duties. Availability percentages should state the observation window and exclusions; planned maintenance and degraded responses can otherwise make comparisons misleading. Resilience comes from redundancy, independent regions and providers, tested Node Recovery, and traffic routing that removes unhealthy instances.

Distinctions and limits

For that reason, availability combines reachability, Node Synchronization, response quality, and role-specific readiness rather than relying only on operating-system uptime. Multiple nodes should not share the same database, network path, or cloud dependency if they are intended to provide fault isolation.

Related definitions: Node Synchronization, Node Recovery.

Example