RPC Error

Overview

How it works

The main risks include retrying permanent errors, hiding provider-specific causes, leaking sensitive data, and assuming every timeout means no action occurred. In JSON-RPC it may include a standardized error object with a code, message, and optional data. When nodes disagree during RPC Error, the application should preserve both observations and defer irreversible action until the authoritative chain state is clear.

In production, teams should define ownership and apply error classification, provider code mapping, request IDs, safe logging, retryability rules, status queries, and alert thresholds. An RPC Error client should validate network identity, units, encoding, and method semantics before acting on provider data.

Useful measures include RPC error rate by method and code, unknown error count, retry success rate, and indeterminate request rate. RPC Error is closely connected to RPC Retry, RPC Timeout, and RPC Monitoring. The RPC Error implementation should distinguish node acceptance, transaction broadcast, execution, confirmation, and indexed visibility where those stages apply.

The RPC Error recovery procedure should distinguish a rejected RPC call from a transaction that was broadcast but not acknowledged. For RPC Error, the client should retain chain, method, block or transaction reference, provider response, and the final on-chain observation.

Provider failover for RPC Error should not create a second transaction or replace a verified chain result with stale node data.

What matters in practice

A remote procedure call can fail at the network, server, protocol or application layer. An HTTP error, JSON-RPC error object and reverted blockchain execution are different outcomes. Classify responses as retryable, permanent or uncertain; for writes, uncertainty is particularly important because a transaction might already have been broadcast. Log request IDs, method, target chain, error code and latency while protecting secrets. Never blindly retry a write with a fresh nonce or transaction just because a node stopped responding.

Related glossary concepts

RPC Retry; RPC Timeout; RPC Monitoring