Storage Architecture¶

Reliable, scalable storage is the foundation of the homelab. TrueNAS Scale provides centralized storage with ZFS for data integrity, iSCSI/NFS for flexibility, and automated snapshots for protection.

TrueNAS Scale¶

Overview¶

TrueNAS Scale is the primary storage platform, serving both block (iSCSI) and file (NFS/SMB) storage to the entire infrastructure.

Key Features: - ZFS File System – Copy-on-write, checksums, data integrity - Storage Pools – Aggregated disks with redundancy (RAID-Z) - Snapshots – Point-in-time recovery with minimal overhead - Replication – Automated sync to backup targets - Apps/VMs – Built-in Kubernetes for containerized apps

Hardware Configuration¶

Disk Array – Multiple pools with different redundancy levels
Hot Spares – Automatic replacement on drive failure
SSD Cache – Read/write cache for performance
10GbE Networking – High-speed connectivity for storage traffic

Storage Types¶

iSCSI (Block Storage)¶

Used for VM disks requiring high performance and low latency.

Use Cases: - Proxmox VM disk storage (shared LUNs) - Database servers (MySQL, PostgreSQL) - High-performance application storage

Configuration:

# iSCSI LUN for Proxmox VMs
Pool: tank-ssd
LUN Size: 500GB
Target: iqn.2024-01.com.truenas:proxmox-vms

Benefits: - Block-level access for direct I/O - Supports live migration of VMs - Compatible with clustering and HA

NFS (Network File Storage)¶

Used for shared file access across multiple systems.

Use Cases: - Media libraries (Jellyfin, Plex) - Backup storage (Proxmox Backup Server) - Container persistent volumes - ISO images and templates

Configuration:

# NFS share for media
Path: /mnt/tank-hdd/media
Permissions: maproot=root
Networks: 10.0.20.0/24, 10.0.40.0/24

Benefits: - Simple file-level sharing - Cross-platform compatibility - Ideal for bulk/archival storage

SMB/CIFS (Windows Shares)¶

Used for Windows clients and workstations.

Use Cases: - User home directories - Shared documents - Windows VM storage

ZFS Features¶

Data Integrity¶

Checksums – Every block verified on read
Self-Healing – Automatic repair with redundancy
Scrubbing – Regular integrity checks

Snapshots¶

Instant Point-in-Time Copies – Minimal space usage
Rollback Capability – Restore to any snapshot
Clone Support – Writable snapshot copies

Snapshot Schedule:

Hourly:  Keep 24 (1 day)
Daily:   Keep 7 (1 week)
Weekly:  Keep 4 (1 month)
Monthly: Keep 12 (1 year)

Replication¶

Automated Sync – Periodic replication to backup server
Incremental Transfers – Only changed blocks sent
Disaster Recovery – Full pool reconstruction from replicas

Backup Strategy¶

3-2-1 Backup Rule¶

3 Copies – Production + local backup + offsite
2 Media Types – Disk + cloud/tape
1 Offsite – Protection against local disasters

Backup Targets¶

Proxmox Backup Server¶

VM/Container Backups – Daily automated backups
Retention – 7 daily, 4 weekly, 6 monthly
Deduplication – Space-efficient incremental backups

TrueNAS Snapshots¶

Local Snapshots – Hourly/daily/weekly/monthly
Instant Recovery – Rollback to any snapshot
Replication – Synced to secondary TrueNAS instance

Cloud Backup¶

Critical Data Only – Vaultwarden, documents, configs
Encrypted Sync – rclone to cloud provider
Offsite Protection – Fire/flood/theft recovery

Storage Pools¶

tank-ssd (Performance Pool)¶

Disks: 4x NVMe SSD in RAID-Z1
Purpose: VM disks, databases, hot data
Performance: High IOPS, low latency

tank-hdd (Capacity Pool)¶

Disks: 8x SATA HDD in RAID-Z2
Purpose: Media, backups, archives
Performance: High throughput, cost-effective

tank-nvme (Cache Pool)¶

Disks: 2x NVMe in mirror
Purpose: ZFS L2ARC (read cache) and SLOG (write cache)
Performance: Accelerates pool access

Monitoring & Alerts¶

Health Monitoring¶

SMART Data – Disk health metrics
Scrub Status – Regular integrity checks
Pool Capacity – Space usage alerts
Replication Status – Backup job monitoring

Alerts¶

Email Notifications – Critical events (disk failure, scrub errors)
Checkmk Integration – Comprehensive monitoring dashboard
Graylog – Storage-related log aggregation

Performance Optimization¶

Caching¶

L2ARC – SSD read cache for frequently accessed data
SLOG – Dedicated SSD for synchronous writes
ARC – RAM cache (automatically managed by ZFS)

Network Optimization¶

Jumbo Frames – MTU 9000 for storage VLAN
10GbE Links – High bandwidth for iSCSI/NFS
LACP Bonding – Link aggregation for redundancy

Tuning¶

# Example ZFS tuning parameters
recordsize=128k      # For media files
compression=lz4      # Transparent compression
atime=off           # Disable access time updates

Best Practices¶

Regular Scrubs – Run monthly ZFS scrubs to detect/repair corruption
Monitor Disk Health – Check SMART data and replace failing disks proactively
Test Restores – Regularly verify backup integrity with test restores
Capacity Planning – Keep pools under 80% usage for optimal performance
Snapshot Retention – Balance space usage with recovery needs
Encryption – Use ZFS native encryption for sensitive data
Documentation – Maintain records of pool configs, shares, and dependencies