Diagnostics, Scrubbing & Recovery Commands

This section details emergency triage workflows, proactive bitrot scrubbing, and Abodi predictive failure scoring using the am CLI.

1. am diagnose — Incident Triage & Forensic Bundles

When troubleshooting degraded clusters or performance bottlenecks:

# Run comprehensive automated incident triage across all cluster layers
am diagnose cluster

# Diagnose specific storage node hardware and network connectivity
am diagnose node storage-node-03

# Generate a cryptographically hashed incident bundle for engineering review
am diagnose report -o /var/log/aarkam/incident-$(date +%s).json --format json

2. am repair — Self-Healing & Bitrot Scrubbing

Aarkam uses continuous background data verification to detect and correct silent bitrot:

# Initiate cluster-wide background data scrub
am repair scrub start --pool sovereign-ec83 --throttle 200MB/s

# View active scrubbing progress, audited chunks, and corrected errors
am repair status

# Reconstruct a specific damaged object from surviving Cauchy parity blocks
am repair reconstruct financial-ledger-2026 "audit/2026-q3.enc" --priority high

3. am abodi — S.M.A.R.T. Scoring & Predictive Health

Query the Abodi diagnostic engine directly from the command line:

# Display health score (0–100) and failure probability for all cluster drives
am abodi scores --format table

# Inspect detailed S.M.A.R.T. counter telemetry for a specific drive
am abodi inspect storage-node-01 /dev/nvme0n1

# View drives predicted to fail within the next 30 days
am abodi alerts --severity warning