Aarkam / Aarkam Wiki / AM CLI / Storage Fleet, Nodes & Volumes Reference ARK-CLI-0303
Aarkam - SAT (Storage Substrate) Fleet & Storage AM CLI • ARK-CLI-0303

Storage Fleet, Nodes & Volumes Reference

Authoritative command reference for am node (Zero-Trust Enrollment, Fencing, Draining), am volume (Physical Mounts), and am placement (HRW Simulation).

Last updated: Nov 03, 2026 Classification: Sovereign Tier-4 Enterprise
Document Identifier ARK-CLI-0303
Standard Status
Production Standard
Architecture Tier Aarkam - SAT (Storage Substrate)
Release Version V-1.0.0-GA (Production Stable)
Architecture Tier Scope: Aarkam - SAT (Standalone Tier) These commands manage physical storage node hosts running the Abodi daemon and NVMe volumes formatted for Kdouja 66,048-byte sector alignment. They enforce zero-trust cryptographic enrollment (NCA ECC-2:2024), hot-swap disk draining, and Highest Random Weight (HRW) rendezvous placement.

This reference details the full command set for managing storage node lifecycles, cryptographic node onboarding, drive failure isolation (fencing/quarantine), volume mount points, and placement ring simulations.


1. am node — Storage Fleet Lifecycle & Maintenance

1.1 List Storage Fleet (am node list)

Lists all registered storage nodes, IP addresses, states, hardware tiers, used capacity percentage, and free space:

bash
# List all online storage nodes sorted by available space
am node list --state Online --sort-by space

# Output raw JSON payload for automated health alerting
am node list --format json | jq '.[] | select(.usedPercentage > 85)'

1.2 Deep Node Inspection (am node inspect)

Displays hardware specifications, mounted volume paths, and real-time Abodi telemetry (temperature, latency, flash wear):

bash
am node inspect storage-01 --abodi

1.3 Isolation & Hot-Swap Draining

Safely fence erratic nodes from write operations or gracefully evacuate chunks prior to drive maintenance:

bash
# Fence erratic node (read-only for existing chunks, blocked from new writes)
am node quarantine storage-02 --reason "Drive read latency elevated above 200ms"

# Gracefully drain all chunks to healthy peers (throttled at 60 MB/s)
am node drain storage-02 --rate 60

# Clear quarantine and rejoin rehabilitated node to active hash ring
am node rejoin storage-02

2. Zero-Trust Cryptographic Node Enrollment

Compliant with Saudi NCA ECC-2:2024 §2-2 zero-trust onboarding, nodes cannot join the cluster without cryptographically bound enrollment tokens:

2.1 Generate Node Enrollment Token (am node token create)

bash
# Mint single-use token valid for 15 minutes bound to node-prod-042
am node token create \
  --node-id "node-prod-042" \
  --role "storage-node" \
  --ttl 15 \
  --coordinator "https://rokka-coordinator.rokka.internal:57774"

2.2 Review and Approve Pending CSRs

bash
# List pending Certificate Signing Requests (CSRs)
am node list-pending

# Sign and issue cluster mTLS identity certificate
am node approve enr_2b6521570e534f64a8b65a75994eeb9f

# Emergency node revocation (broadcasts revocation across Raft consensus)
am node revoke storage-compromised-03 --reason "Hardware tamper detection triggered"

3. am volume — Physical Disk Mounts & Direct I/O

Inspects mounted storage partitions and raw block devices hosting Kdouja PageStore segments:

bash
# List mounted volumes across all storage hosts
am volume list

# Inspect volume 1 on storage-01 (mount path, Kdouja WAL segment directory)
am volume inspect storage-01 1

4. am placement — HRW Hash Ring & Domain Simulation

Evaluates Highest Random Weight (HRW) rendezvous distribution across racks and failure domains:

bash
# Simulate HRW distribution for 1,000,000 keys across active topology
am placement simulate --keys 1000000

# Test placement target nodes for a specific object key
am placement test "backups/2026/db-dump.tar.gz"