Chain, RPC and transaction troubleshooting registry¶
Use these stable troubleshooting IDs for chain identity, RPC transport/method behavior, transaction submission, nonce/gas/replacement issues, receipt interpretation, finality and reorganization symptoms. Canonical chain/node state outranks gateway, Explorer, Indexer, Search or application presentation.
A timeout is not proof that a write failed. Before retrying any signed/state-changing action, preserve the transaction hash/raw transaction identity where available and inspect canonical state.
TRB-CHAIN-001 — Network or chain identity does not match¶
- Audience: user, developer, operator
- Surface: wallet, dApp, RPC client, deployment tooling
- Symptom: the reported chain ID/environment differs from the intended 420 Integrated network, or configured contracts/services do not match the selected network.
- Severity: blocked
- Authority source: approved network manifest/chain identity plus canonical RPC
eth_chainId/network evidence. - Likely causes: wrong endpoint; stale/manual network configuration; local/devnet values promoted into another environment; application connected to a different wallet network.
- Retry safety: not-applicable until identity is corrected.
- Safe diagnostics: endpoint label, chain ID, environment name, client version, approved manifest identity. Do not share endpoint credentials.
- Recovery: stop writes; compare chain ID and approved network manifest; switch to a qualified endpoint; re-resolve deployment/service identities for that environment; reconnect clients.
- Escalate when: approved identity and canonical RPC disagree or the canonical environment cannot be established.
TRB-RPC-001 — RPC endpoint unavailable or transport failing¶
- Audience: user, developer, operator
- Surface: public RPC/WSS, gateway/proxy, client transport
- Symptom: connection refused, timeout, TLS/proxy failure, repeated disconnects or no JSON-RPC response.
- Severity: degraded
- Authority source: endpoint health plus direct qualified canonical-node RPC where available.
- Likely causes: gateway outage; rate limiting; network path/TLS failure; stale endpoint; node process unavailable.
- Retry safety: safe for read-only calls; conditional for writes.
- Safe diagnostics: endpoint hostname, timestamp, method name, HTTP/WebSocket status, sanitized error, health/readiness response.
- Recovery: distinguish transport failure from method error; test a read-only identity/health request; fail over to another qualified endpoint; if operating the service, verify proxy then node health.
- Stop before write retry: confirm whether the signed transaction was already accepted by any endpoint or visible by hash.
- Escalate when: multiple qualified endpoints fail or canonical node health is unavailable.
TRB-RPC-002 — RPC method rejected, invalid or unsupported¶
- Audience: developer, operator
- Surface: public JSON-RPC
- Symptom:
method not found, invalid request/params, policy rejection, or unsupported public method. - Severity: blocked
- Authority source: generated public RPC reference and runtime RPC response.
- Likely causes: private/admin method used on public RPC; malformed params; unsupported compatibility assumption; client version mismatch.
- Retry safety: safe only after correcting the request.
- Safe diagnostics: method name, sanitized params shape, JSON-RPC error code/message, client build.
- Recovery: verify the method is in the public RPC surface; correct request/parameter shape; do not substitute private Engine/admin/signer methods through a public gateway.
- Escalate when: a documented public method is rejected consistently by a healthy qualified endpoint.
TRB-RPC-003 — RPC data disagrees between endpoints¶
- Audience: developer, operator
- Surface: multiple RPC providers/gateways
- Symptom: block number, receipt, balance, nonce or logs differ across endpoints.
- Severity: degraded
- Authority source: canonical node state at the relevant block/finality level.
- Likely causes: provider lag; different head/safe/finalized views; reorg; stale cache; endpoint on wrong chain.
- Retry safety: conditional.
- Safe diagnostics: chain ID, block hash/number, finality tag, endpoint identities, transaction hash.
- Recovery: verify all endpoints are on the same chain; compare explicit block hashes/finality tags instead of only
latest; prefer canonical/finalized evidence for irreversible decisions. - Escalate when: finalized-state disagreement persists across qualified canonical sources.
TRB-TX-001 — Transaction submission timed out or returned an ambiguous result¶
- Audience: user, developer
- Surface: wallet, SDK, RPC
eth_sendRawTransaction - Symptom: submission timed out/disconnected and the caller does not know whether the transaction was accepted.
- Severity: value-risk
- Authority source: canonical transaction lookup/receipt plus sender nonce state.
- Likely causes: response lost after acceptance; gateway timeout; endpoint failover; transient node pressure.
- Retry safety: unsafe blind retry.
- Safe diagnostics: transaction hash if locally derivable/returned, sender, nonce, chain ID, submission timestamp.
- Recovery: preserve the signed transaction identity; query by hash; inspect sender nonce/pending state; only resubmit the same signed transaction when the client flow explicitly treats that as safe; do not create a fresh value-moving transaction merely because the response timed out.
- Escalate when: canonical lookup and nonce state remain ambiguous.
TRB-TX-002 — Nonce too low, nonce too high or nonce lane blocked¶
- Audience: user, developer
- Surface: EOA/smart-account transaction submission
- Symptom: nonce-related rejection, later transactions remain pending behind an earlier nonce, or local nonce differs from canonical state.
- Severity: blocked
- Authority source: canonical account nonce plus known pending/included transactions.
- Likely causes: concurrent senders; stale local nonce cache; earlier pending transaction; already-included/replaced transaction.
- Retry safety: conditional.
- Safe diagnostics: sender, chain ID, candidate nonce, canonical transaction count, known hashes for nearby nonces.
- Recovery: rebuild nonce view from canonical state and known pending transactions; resolve the earliest missing/pending nonce first; use explicit replacement/cancellation semantics if supported.
- Escalate when: signing infrastructure is issuing conflicting transactions for the same nonce without an intentional replacement policy.
TRB-TX-003 — Insufficient balance, fee or gas condition¶
- Audience: user, developer
- Surface: transaction preparation/submission
- Symptom: insufficient funds, intrinsic gas/fee rejection, estimate failure tied to balance/fee constraints.
- Severity: blocked
- Authority source: canonical balance, fee data and execution rules.
- Likely causes: insufficient
$420for value plus gas; stale balance; incorrect gas/fee assumptions; state changed since simulation. - Retry safety: conditional after correcting funding/fee inputs.
- Safe diagnostics: public sender, value, gas limit, fee parameters, canonical balance, estimate result.
- Recovery: refresh canonical balance/fee data; re-estimate; reduce/adjust the intended action only when semantically valid; never bypass a revert/simulation failure by arbitrarily raising gas.
- Escalate when: estimates fail despite valid balance and a known-good call against current state.
TRB-TX-004 — Replacement or cancellation is not taking effect¶
- Audience: user, developer
- Surface: pending transaction replacement
- Symptom: replacement remains pending, original still appears, or multiple same-nonce candidates are visible.
- Severity: value-risk
- Authority source: canonical inclusion state for the nonce and transaction hashes.
- Likely causes: replacement fee insufficient; original already included; provider mempool differences; replacement not propagated.
- Retry safety: unsafe without nonce/hash inspection.
- Safe diagnostics: sender, nonce, all candidate transaction hashes, fee parameters, inclusion/receipt status.
- Recovery: identify whether any candidate is already included; if none are included and replacement is supported, follow wallet/client replacement policy; do not keep generating same-nonce variants blindly.
- Escalate when: conflicting signed transactions exist unexpectedly or value destinations differ.
TRB-TX-005 — Transaction is pending longer than expected¶
- Audience: user, developer
- Surface: transaction lifecycle
- Symptom: a known hash has no receipt and remains pending/unconfirmed.
- Severity: degraded
- Authority source: canonical transaction/nonce state and current chain progress.
- Likely causes: fee competitiveness; blocked nonce lane; temporary congestion; propagation issue; endpoint lag.
- Retry safety: conditional.
- Safe diagnostics: hash, sender, nonce, fee parameters, chain head progress, earlier nonce transactions.
- Recovery: verify the chain is progressing; resolve earlier nonce blockers; use explicit replacement only if supported; otherwise wait rather than duplicating the semantic action.
- Escalate when: the transaction disappears from all qualified sources while nonce state remains unresolved.
TRB-TX-006 — Included transaction failed/reverted¶
- Audience: user, developer
- Surface: execution receipt
- Symptom: receipt exists but execution status indicates failure/revert.
- Severity: blocked
- Authority source: canonical receipt and runtime revert/error evidence.
- Likely causes: changed contract state; insufficient authority; expired deadline; invalid calldata; protocol precondition failure.
- Retry safety: unsafe until the cause is understood; gas may already have been consumed.
- Safe diagnostics: transaction hash, receipt, block/finality status, surfaced revert/error identifier, target address.
- Recovery: inspect receipt/runtime error; re-read canonical state; correct the underlying precondition; re-simulate before any new signed attempt.
- Escalate when: the same valid call consistently fails without a documented/runtime-identifiable cause.
TRB-CHAIN-002 — Receipt/block disappeared or changed after a reorganization¶
- Audience: user, developer, operator
- Surface: blocks, receipts, logs, application confirmation state
- Symptom: a previously seen receipt/block/log is missing or replaced, or transaction confirmation count decreases.
- Severity: degraded
- Authority source: current canonical head plus safe/finalized mapping.
- Likely causes: normal pre-finality reorganization; provider lag; derived index not yet reconciled.
- Retry safety: conditional.
- Safe diagnostics: transaction hash, old/new block hash, block number, prior/current finality state.
- Recovery: re-query canonical transaction/receipt state; wait for the required finality level; allow Indexer/Explorer projections to reconcile; do not mutate canonical state merely to make a derived view match.
- Escalate when: a transaction previously reported finalized is no longer canonical.
TRB-CHAIN-003 — Head, safe and finalized states are being confused¶
- Audience: user, developer, operator
- Surface: confirmation/finality handling
- Symptom: an application treats
latestas irreversible, or different interfaces show different confirmation states. - Severity: value-risk
- Authority source: consensus-defined head/safe/finalized mapping and canonical RPC tags.
- Likely causes: confirmation policy too weak; UI collapses finality states; provider only showing head progress.
- Retry safety: not-applicable.
- Safe diagnostics: block hash/number and its head/safe/finalized status.
- Recovery: apply the task-appropriate finality requirement; for value/cross-chain/irreversible decisions, do not substitute mere inclusion for the required safe/finalized state.
- Escalate when: canonical endpoints disagree on finalized mapping.
TRB-CHAIN-004 — Explorer/Indexer presentation disagrees with canonical RPC¶
- Audience: user, developer, operator
- Surface: Explorer, Indexer, Search, application activity views
- Symptom: balance, receipt, log, status or block presentation differs from canonical RPC.
- Severity: degraded
- Authority source: canonical chain/protocol state; derived services remain rebuildable/non-authoritative.
- Likely causes: indexing lag; reorg reconciliation; stale cache; service replay/rebuild.
- Retry safety: unsafe if the only reason to retry a write is stale derived presentation.
- Safe diagnostics: transaction/block hash, canonical RPC result, derived service cursor/status, timestamps.
- Recovery: trust canonical state for action decisions; allow/rebuild the derived service; cross-link the discrepancy into Indexer/Explorer troubleshooting rather than resubmitting chain writes.
- Escalate when: derived state fails to converge after canonical finality and normal recovery/replay.