Ejecuta cualquier Skill en Manus
con un clic

Ejecuta cualquier Skill en Manus con un clic

fukuii-log-triage

Estrellas3

Forks3

Actualizado30 de junio de 2026 a las 18:22

Configure Fukuii logging and triage problems through logs — set the log level at runtime without a restart, switch JSON output, find the log files, and map common error patterns (peer drops, RocksDB, OOM, sync stalls) to causes. Use when investigating any node issue via logs, when asked to raise/lower verbosity, or to enable structured logging. Reading logs is read-only; changing the live log level is a reversible write under the guarded-write protocol.

Instalación

Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.

Ejecutar en Manus

Fuente

chippr-robotics

chippr-robotics/fukuii

Abrir repositorio de GitHub Ver repositorios del creador

Descarga

Ejecutar en Manus

Ocupaciones relacionadasSOC

Basado en la clasificación ocupacional SOC

Administradores de redes y sistemas informáticosOcupaciones informáticas y matemáticas·SOC 15-1244

SKILL.md

readonly

Más de este repositorio

mismo repositorio

fukuii-checkpoint-service

chippr-robotics/fukuii

Operate Fukuii's checkpoint-based rapid sync for ETC/Mordor nodes — point a fresh node at a trusted checkpoint archive (local file or HTTP URL) so it imports a known pivot and starts SNAP sync immediately instead of waiting on peer consensus, and produce checkpoint archives for others. Use when bootstrapping an ETC/Mordor node fast, running a checkpoint source for a fleet, or debugging why checkpoint import didn't engage. ETH/Sepolia nodes do NOT use checkpoint-sync — they sync via Consensus Layer / Engine API instead. Config changes require a restart — an irreversible/disruptive action under the guarded-write protocol.

2026-06-303

fukuii-cl-setup

chippr-robotics/fukuii

Stand up a Consensus Layer (CL) client (Prysm / Lighthouse / Teku) alongside Fukuii for an ETH/Sepolia node so the CL can drive Execution Layer block import via the Engine API. Use when bringing up a post-Merge ETH/Sepolia node, connecting or replacing the CL, or when the operator asks "why won't my Sepolia node sync" / "do I need a beacon client". Covers the EL+CL pairing rationale, JWT secret sharing, CL connection flags, startup order, and the 401 liveness check. ETH/Sepolia only — ETC/Mordor has no CL. For the EL-side Engine API config keys see fukuii-engine-api-setup; for handshake/auth faults see fukuii-engine-api-debug.

2026-06-303

fukuii-disk-management

chippr-robotics/fukuii

Triage and manage Fukuii's disk usage — understand the datadir layout, measure RocksDB growth, free space safely, and size storage for the supported chains. Use when disk is filling up, the node warns about space, you're planning capacity, or sync is slow due to I/O. Measurement is read-only; any deletion/relocation of the database is an irreversible action under the guarded-write protocol.

2026-06-303

fukuii-engine-api-debug

chippr-robotics/fukuii

Troubleshoot a broken or stalled Engine API channel between a Consensus Layer client and Fukuii on ETH/Sepolia — JWT 401 auth failures, forkchoiceUpdated timeouts, newPayload INVALID, EL-reports-SYNCING, exchangeCapabilities version mismatches — and decide whether the fault is in the EL (Fukuii) or the CL. Use when the CL logs Engine API errors, payloads are rejected, attestation duties are withheld, or sync won't advance with a connected CL. ETH/Sepolia only; ETC/Mordor has no Engine API. For first-time CL bring-up see fukuii-cl-setup; for EL config keys see fukuii-engine-api-setup.

2026-06-303

fukuii-engine-api-setup

chippr-robotics/fukuii

Configure Fukuii's Engine API authrpc endpoint so a Consensus Layer (CL) client can drive ETH/Sepolia block import — generate the shared JWT secret, enable the authrpc server, and verify the CL handshake. ETH/Sepolia only; ETC/Mordor does not use the Engine API. Use when setting up a new ETH/Sepolia node, connecting or replacing the CL, or troubleshooting eth_syncing not advancing when peers are healthy. Config changes require a restart — irreversible/disruptive under the guarded-write protocol.

2026-06-303

fukuii-first-start

chippr-robotics/fukuii

Bring up a brand-new Fukuii node end to end — pick the network, generate the node key, set the datadir and config, start the node, and confirm it discovers peers and begins syncing. Use when bootstrapping a fresh node, onboarding a new host, or setting up ETC/Mordor/Sepolia for the first time. For ETH/Sepolia nodes a Consensus Layer client must also be running and connected to the Engine API before the EL will sync. Generating keys and starting the node are security-sensitive/irreversible actions under the guarded-write protocol.

2026-06-303

name

fukuii-log-triage

description

Fukuii log triage

Read ../CONVENTIONS.md first. Reading logs is 🟢; admin_changeLogLevel is 🟡 (state the level and confirm — DEBUG/TRACE are verbose and can fill disk).

When to use

Diagnosing almost any node problem (logs are the primary signal).
Temporarily raising verbosity to capture a transient issue, then lowering it.
Turning on JSON logs for aggregation (Loki/ELK).

Log locations & config

Files: ~/.fukuii/<network>/logs/fukuii.log (rotated; *.log.zip, 10 MB × up to 50). Configured by src/main/resources/logback.xml.
Levels: TRACE < DEBUG < INFO (default) < WARN < ERROR.
Static config (logging { logs-level, json-output, logs-dir, logs-file }) or env FUKUII_LOG_LEVEL / FUKUII_LOGGING_JSON_OUTPUT=true — both need a restart.

Procedure

Set level at runtime (🟡, no restart) — admin_changeLogLevel("<LEVEL>"). Use this to bump to DEBUG just long enough to capture the symptom, then return to INFO. Confirm before going to DEBUG/TRACE (volume + disk).
Read & scan (🟢) — tail the live file; grep for ERROR/WARN, stack traces, and the patterns below around the incident window.
Classify (🟢) — map to the table, then hand off.

Pattern → cause → hand-off

Log pattern	Likely cause	Hand off to
Repeated peer disconnect / handshake fail	P2P/connectivity	`fukuii-peer-management`
`AccountRange`/pivot stall, snap errors	ETC sync stall	`fukuii-sync-troubleshooting`
Engine API timeout / missing `engine_forkchoiceUpdated`	ETH/Sepolia: CL not connected or stalled	`fukuii-sync-troubleshooting` — verify CL is running
RocksDB corruption / IO error	disk/DB	`fukuii-disk-management`, then restore
`OutOfMemoryError`, GC thrash	heap/resources	tune JVM heap (`.jvmopts`); restart
`No space left on device`	disk full	`fukuii-disk-management`
TLS/cert errors on RPC	RPC TLS	`fukuii-tls-operations`

Deep reference

docs/runbooks/log-triage.md (pattern catalogue, analysis tooling)
docs/operations/LOGGING.md, docs/operations/metrics-and-monitoring.md

Output

CONVENTIONS §4 block. Quote the specific log lines as evidence; record any level change made and that it was reverted.