Operations, Maintenance & Disaster Recovery Reference
Authoritative command reference for am maintenance (Servicing Windows), am rebalance, am snapshot, am backup, am restore, am cert, and am job.
This reference details production workflows for servicing live hardware without taking the storage cluster offline, capturing point-in-time snapshots, rotating PKI identities, and tracking asynchronous cluster jobs.
1. am maintenance — Safe Node Servicing Windows
Orchestrates graceful maintenance windows, preventing split-brain or premature shard re-silvering during scheduled hardware repairs:
# Place node into maintenance for 2 hours with audit justification
am maintenance start storage-03 --duration 2h --reason "NVMe Drive Replacement"
# List all active maintenance windows across cluster
am maintenance list
# Complete maintenance and restore node to full active read/write status
am maintenance stop storage-03
2. am rebalance — Dynamic Capacity Balancing
Re-distributes Cauchy chunks evenly across storage nodes following hardware expansions:
# Preview planned chunk rebalancing movements without mutating state
am rebalance start --dry-run
# Launch background rebalance throttled at 100 MB/s
am rebalance start --max-rate 100
# Monitor live rebalance migration progress
am rebalance status --watch 3s
# Temporarily pause rebalancing during heavy user ingest
am rebalance pause
3. Snapshots, Backups & Restoration
3.1 Point-in-Time Snapshots (am snapshot)
# Create atomic point-in-time snapshot of volume 1
am snapshot create 1 --name "snap-pre-upgrade-2026-q3"
# List snapshots with creation timestamps and referenced sizes
am snapshot list
# Rollback volume to snapshot state in an emergency
am snapshot rollback snap_a7b8c9d0e1f2 --yes
3.2 Off-Cluster Cold Backups & Restoration (am backup & am restore)
# Export encrypted off-cluster backup bundle with Brotli compression
am backup create --target s3://remote-cold-archive/ --compression brotli
# Check status of backup worker
am backup status
# Launch asynchronous volume restoration from backup archive
am restore start bkp_9876543210 --target-volume 1
am restore status --watch 2s
4. am cert — PKI Identity & Zero-Downtime Rotation
Compliant with NCA ECC-2:2024 §2-2-4 mandatory cryptographic rotation, rotates internal mTLS certificates without dropping active gRPC streams:
# List all active node certificates with expiration dates and SANs
am cert list
# Inspect certificate details for storage-01
am cert inspect storage-01
# Trigger automated zero-downtime mTLS certificate rotation for a node
am cert rotate storage-01 --yes
5. am job — Universal Asynchronous Job Tracking
Universal job control plane managing rebalances, disk evacuations, scrubbing, and restorations:
# List running background jobs
am job list --state Running
# Inspect deep execution telemetry for a job
am job inspect job-f956395e506c
# Stream the last 25 lines of execution logs for a job
am job logs job-f956395e506c --tail 25
# Abort a running background job cleanly
am job cancel job-f956395e506c