troubleshoot-zfs

Troubleshoot ZFS pool degradation by querying Netdata metrics via MCP to identify failure archetypes like fragmentation collapse, TXG sync hangs, or ARC memory starvation. It follows an operator playbook decision tree to route diagnostics and verify remediation.

ZFS write freezes and fsync latency spikes are often caused by TXG sync hangs or ZIL/SLOG bottlenecks. This skill diagnoses these incidents by analyzing Netdata-observed I/O latency, queue depth, and internal ZFS state metrics to verify the root cause.

Detect silent ZFS data corruption and ARC memory starvation by analyzing checksum errors and system memory pressure signals. This skill queries Netdata MCP for ZFS availability, errors, and resource utilization to confirm if ARC size is causing OOM events.

You can verify ZFS pool health and remediation success using Netdata by re-running the same MCP queries. This confirms per-vdev state, errors, and throughput return to expected ranges rather than stopping at the first detected anomaly.

Diagnose ZFS fragmentation-driven write amplification collapse by evaluating capacity limits and throughput metrics. This skill uses Netdata signals to verify if fragmentation has pushed the pool over a performance cliff and routes you through the remediation path.