Hard Drive Health Report — HP3Proxmox#

This report is a reference for selecting disks for non-critical storage. Disks are identified by model and serial number; Linux device names such as /dev/sdb are intentionally omitted because they can change.

Quick reference#

The complete inventory table is maintained in the Homelab - Storage inventory Google Sheet.

1 Spares#

1.1 Seagate ST4000DM000-1F2168 — Z300VMNK#

• Capacity: 4 TB
• Test date: 2026-09-16
• Power-on time: 37,294 hours
• SMART health: PASSED
• Extended self-test: Completed without error
• SMART error log: No errors logged
• Reallocated sectors: 0
• Current pending sectors: 0
• Offline uncorrectable sectors: 0
• Reported uncorrectable errors: 0
• Runtime bad blocks: 1
• UDMA CRC errors: 2
• Load-cycle count: 316,566 raw cycles; normalized value at minimum (1)

Assessment: The extended test completed successfully and there are no pending, reallocated, or offline-uncorrectable sectors. The disk is old and heavily used, with one runtime bad-block event, two interface CRC errors, and an extremely high load-cycle count.

Home Assistant Database Management Project Plan#

First establish safe Recorder purge and repack settings. Then create and validate an export target for historical Home Assistant data.

Keep approximately 30 days of detailed history locally. Complete and verify the Recorder maintenance setup before starting the export work.

1. Configure Recorder Purge and Repack#

• 1.1 Set purge_keep_days to 30.
• 1.2 Enable auto_purge.
• 1.3 Enable auto_repack.
• 1.4 Validate the configuration and restart Home Assistant if required.
• 1.5 Confirm that automatic purge and repack are active.
• 1.6 Confirm that the database remains within the 30-day retention period.
• 1.7 Confirm that sufficient free space remains for maintenance operations.

Do not begin the export-target work until the Recorder purge and repack settings have been configured and verified.

Backup Strategies#

1 Backrest & Restic Google Drive Backup#

2026-09-01 Not used anymore

Encrypts TrueNAS/DSM network shares with Restic and stores them in Google Drive through Rclone. See Backrest & Restic Google Drive Backup Setup.

2 Proxmox PBS Backup to Google Drive#

2026-09-01 Not used anymore

Synchronizes the deduplicated PBS datastore from the UNAS-Pro NAS to Google Drive through MMDocker and Rclone. See Proxmox - PBS Backups via NAS with to GDrive.

Synology Administration Guide#

1 Syncthing Deployment#

1.1 Synology DSM Implementation (Docker)#

1. Inotify Limit:

   echo "fs.inotify.max_user_watches=204800" | sudo tee -a /etc/sysctl.d/90-override.conf
   sudo sysctl -p /etc/sysctl.d/90-override.conf

2. User IDs: Use Task Scheduler to run id > /volume1/docker/id.txt to find UID/GID (e.g., 1027/100).
3. Docker Compose Highlights:

   environment:
     - PUID=1027
     - PGID=100
     - TZ=America/New_York
   volumes:
     - ./config:/config
     - /volume2/HP3Storage:/volume2/HP3Storage

1.2 Standard .stignore Patterns#

Create a .stignore file in the root of the sync folder:

(?i)#recycle
@eaDir
.DS_Store
.stfolder
.stignore
$RECYCLE.BIN
Thumbs.db
Desktop.ini
.SynologyWorkingDirectory

2 Storage & Maintenance Commands#

2.1 iPerf3#

sudo -i
synogear install
iperf3 -s -p 11111

2.2 Locked Folder Cleanup#

Remove Synology metadata directories when cleaning up a locked folder:

Master Plan: UNAS Pro SMB Usage Inventory#

This document outlines the proposed step-by-step strategy to perform a comprehensive audit and inventory of all workloads, computers, virtual machines, Docker stacks, and scripts utilizing SMB shares from the UNAS Pro (10.1.2.2) NAS.

HP6_UNAS_Pull_Backup.sh is the current backup script for HP6TrueNAS (10.1.1.16). HP1Cloud is the current name for VM 217 at 10.1.2.219. HP8_UNAS_Pull_Backup.sh is retained as the legacy filename for the active HP7TrueNAS pull job.

UniFi UNAS-Pro Pool Migration Strategy: Pool 1 (SSD) to Pool 2 (HDD)#

This document defines the operational plan, safety protocols, database update queries, and verification steps to migrate all shared folders from Storage Pool 1 (SSD - Degraded RAID5) to Storage Pool 2 (HDD - Healthy RAID5) on the master UniFi UNAS-Pro NAS (10.1.2.2) while preserving all access rights, Samba configurations, and TrueNAS backup operations.

The checked commands in this historical migration record now use the current identity HP6TrueNAS at 10.1.1.16; the node was formerly HP7TrueNAS at 10.1.1.17.

UniFi UNAS-Pro Master Backup Strategy: Native Rsync Daemon (Pull) Architecture#

This document defines the complete operational standard, architecture, schedule matrix, node-by-node configuration reference, and disaster recovery procedures for backing up all shared data from the master UniFi UNAS-Pro NAS (10.1.2.2) across the fleet of 5 TrueNAS SCALE storage nodes (HP1, HP2, HP3, HP6, HP7). It is designed to be sufficient to fully reconstruct the entire setup from scratch.

The former HP7TrueNAS node (10.1.1.17) is now HP6TrueNAS. The storage dataset, SMB shares, and backup script have been renamed to the HP6 identifiers.

UNAS-Pro Backup Migration Plan: SMB (Push) to Native Rsync Daemon (Pull)#

Antigravity Chat: agy –conversation=4c0c9b6d-2722-4fd6-a4aa-638c75ab46f4

This document is the complete master migration plan for transitioning the UNAS-Pro backup strategy from the UniFi Drive UI-managed SMB (Push) setup to a Native Rsync Daemon (Pull) model.

Because the UniFi Drive UI natively supports Rsync server settings and shared folder modules, migrating to Rsync requires configuring your TrueNAS SCALE servers (HP1, HP2, HP3, HP6, HP7) to pull the backups directly from the UNAS-Pro Rsync daemon over TCP port 873.

Proxmox Backup Server - Mounting an SMB NAS Share as a PBS Datastore#

This note outlines the step-by-step process used to mount a remote SMB/Samba share from a NAS onto a Proxmox Backup Server (PBS) host at the OS level, assign proper permissions for the PBS daemon, and initialize it as a new backup datastore.

Prerequisites#

• A target share on the NAS (e.g., //10.1.2.2/ProxmoxBackups).
• An existing SMB credentials file on the PBS host (e.g., /root/.smb_credentials).

Step-by-Step Implementation#

Step 1: Create the Local Mount Directory#

Create a dedicated folder at the root of the PBS host filesystem where the network share will be attached.

NAS Backup Pipeline: Backrest, Restic & Google Drive#

Complete configuration guide for backing up isolated TrueNAS/DSM network shares from the headless Ubuntu VM (HP1Docker) to an encrypted Google Drive repository using Backrest (a Web UI wrapper for Restic) and Rclone. For baseline Rclone installation, remote management, and CLI commands, see the RClone Guide.

1. System Architecture Overview#

• Source Data: 8 explicit network shares mounted via CIFS/SMB on the host VM at /mnt/<ShareName>.
• Backup Host: Ubuntu VM (HP1Docker - 10.1.2.202).
• Backup Agent: Backrest Docker Container running Restic 0.18.1.
• Target Backend: Google Drive (proxmox.app@gmail.com) via a dedicated Rclone remote.
• Security Model: Client-side, zero-knowledge encryption via a dedicated repository password before data leaves the local host.
• Rate-Limit Protections: Powered by a custom Google Cloud Project ID to bypass global multi-user API query restrictions during multi-terabyte data transfers.

2. Directory Structure Setup#

All configuration and container runtime files are centralized on the host under the user deployment directory.