Skip to content

03 - USB HDD + SnapRAID Sync

Date: 2025-12-22/23
System: Proxmox VE + Docker LXC (ID: 100, IP: 192.168.0.110)

📋 Starting State

  • Proxmox server: Working MergerFS + SnapRAID
  • disk1 (5.5TB HGST) - data
  • disk2 (5.5TB HGST) - parity
  • Raspberry Pi server: DEAD ❌
  • 2 USB HDDs connected:
  • disk3 (1TB) - "Filmek" label
  • disk4 (2TB) - "Filmek2" label
  • Docker LXC: Running, but media paths use old structure

🎯 Goals

  1. ✅ Integrate USB HDDs into the MergerFS pool
  2. ✅ Update SnapRAID config (3 data disks)
  3. ✅ Unify media structure (/media/movies/hun, /media/tv/eng)
  4. ✅ Fix Docker application paths
  5. ✅ Resolve DNS issue (RPi AdGuard Home died)

1️⃣ Identifying and Mounting USB HDDs

Check disks:

lsblk
blkid /dev/sdc1
blkid /dev/sdd1

Result: 

sdc1: 931GB, ext4, Label="Filmek"
      UUID=YOUR_DISK3_UUID
sdd1: 1.8TB, ext4, Label="Filmek2"
      UUID=YOUR_DISK4_UUID

Mount points and mounting:

mkdir -p /mnt/disk3 /mnt/disk4
mount /dev/sdc1 /mnt/disk3
mount /dev/sdd1 /mnt/disk4
df -h | grep disk

Data: - disk3: 713GB used (82% full) - disk4: 922GB used (53% full)

fstab configuration (automount):

nano /etc/fstab

Lines added: 

# USB HDDs (from RPi)
UUID=YOUR_DISK3_UUID  /mnt/disk3  ext4  defaults,noatime  0  2
UUID=YOUR_DISK4_UUID  /mnt/disk4  ext4  defaults,noatime  0  2

Test: 

umount /mnt/disk3 /mnt/disk4
mount -a
df -h | grep disk

2️⃣ Expanding the MergerFS Pool

Original MergerFS config:

/mnt/disk1 -> /mnt/storage

New MergerFS config (3 disk pool):

nano /etc/fstab

Modified line: 

/mnt/disk1:/mnt/disk3:/mnt/disk4  /mnt/storage  fuse.mergerfs  defaults,allow_other,use_ino,cache.files=partial,dropcacheonclose=true,category.create=mfs  0  0

Remount:

umount /mnt/storage
systemctl daemon-reload
mount /mnt/storage
df -h /mnt/storage

Result: 

1:3:4    8.1T  1.7T  6.1T  21% /mnt/storage

✅ 8.1TB unified pool (5.5T + 1T + 2T)

3️⃣ Updating SnapRAID Configuration

Edit SnapRAID config:

nano /etc/snapraid.conf

New configuration: 

# SnapRAID Configuration

# Parity file (disk2 = 5.5TB parity)
parity /mnt/disk2/snapraid.parity

# Content files (saved in multiple locations!)
content /var/snapraid.content
content /mnt/disk1/.snapraid.content
content /mnt/disk3/.snapraid.content
content /mnt/disk4/.snapraid.content

# Data disks (3 disks in the pool)
data d1 /mnt/disk1
data d3 /mnt/disk3
data d4 /mnt/disk4

# Exclude patterns
exclude *.unrecoverable
exclude /tmp/
exclude /lost+found/
exclude *.!sync
exclude *.tmp

# Autosave
autosave 500

Run SnapRAID sync:

snapraid status  # Check before sync
snapraid sync    # Generate parity

Sync result: 

Scanned d1 in 0 seconds   ✅
Scanned d3 in 3 seconds   ✅
Scanned d4 in 14 seconds  ✅

100% completed, 1766832 MB (1.7TB)
Duration: 2:43 (163 minutes)
Average speed: 165-239 MB/s

Everything OK ✅

4️⃣ Restructuring the Media Directory

Problem:

  • disk3/disk4: Movies were directly under /movies/ and /tv/
  • Goal: /media/movies/hun, /media/movies/eng, /media/tv/hun, /media/tv/eng

Restructure (disk3):

mkdir -p /mnt/disk3/media
mv /mnt/disk3/movies /mnt/disk3/media/
mv /mnt/disk3/tv /mnt/disk3/media/
ls -lh /mnt/disk3/media/

Restructure (disk4):

mkdir -p /mnt/disk4/media
mv /mnt/disk4/movies /mnt/disk4/media/
mv /mnt/disk4/tv /mnt/disk4/media/
ls -lh /mnt/disk4/media/

Result: 

disk1/media/movies/hun/  ← EMPTY (new)
disk1/media/movies/eng/  ← EMPTY (new)
disk3/media/movies/hun/  ← 262GB movies (144 folders)
disk3/media/movies/eng/  ← Movies (7 folders)
disk4/media/movies/hun/  ← 231GB movies (100 folders)
disk4/media/movies/eng/  ← Movies (3 folders)

Permission fix:

chmod -R 777 /mnt/disk3/media/
chmod -R 777 /mnt/disk4/media/

5️⃣ Resolving the MergerFS Merge Problem

Problem:

ls -lh /mnt/storage/media/movies/hun/  # EMPTY! ❌

Reason: Empty hun/eng folders existed on disk1 → MergerFS was only showing those!

Solution:

# Remove empty folders from disk1
rmdir /mnt/disk1/media/movies/hun
rmdir /mnt/disk1/media/movies/eng
# tv/hun remains (Lakásvadászok series)

# Verify
ls -lh /mnt/storage/media/movies/hun/ | head -5

Result: 

total 960K
drwxrwxrwx  2 100000 100000 4.0K  101.Dalmatians.1961...
drwxrwxrwx  3 100000 100000 4.0K  101.Dalmatians.1996...
... (654 Hungarian movies total!)

✅ MergerFS now merges the content of all 3 disks!

6️⃣ Docker LXC Media Visibility Problem

Problem:

# Inside LXC
pct enter 100
ls -lh /mnt/storage/media/movies/hun/  # Only 2 folders! ❌

# On Proxmox host
ls -lh /mnt/storage/media/movies/hun/  # 654 folders! ✅

Reason: The MergerFS mount was not yet ready when the LXC booted!

Solution: LXC restart

# On Proxmox host
pct stop 100
pct start 100
pct enter 100

# Verify
ls -lh /mnt/storage/media/movies/hun/ | wc -l
# Result: 654 ✅

7️⃣ Resolving Jellyfin Problems

Problem 1: Marker file conflict

Error: Expected to find only .jellyfin-data but found marker for /config/.jellyfin-config

Solution: 

rm /srv/docker-data/jellyfin/.jellyfin-config
docker start jellyfin

Problem 2: Media not visible

docker exec -it jellyfin ls -lh /media/hun/
# Result: total 0  ❌

Reason: Container was still using the old path mapping!

Solution: 

cd /srv/docker-compose/jellyfin
docker compose down
docker compose up -d
docker exec -it jellyfin ls -lh /media2/hun/

Result: 

total 112K
drwxrwxrwx   2 root root 4.0K  Ahsoka.S01...
drwxrwxrwx  12 root root 4.0K  A.Kiraly.S01...
... (28+ series!) ✅

8️⃣ Resolving DNS Problem

Problem:

# Inside LXC
ping google.com  # Timeout ❌
ping 8.8.8.8     # Works ✅
cat /etc/resolv.conf

Result: 

nameserver 192.168.0.102  ← OLD RPi AdGuard Home!

Temporary fix (inside LXC):

nano /etc/resolv.conf

nameserver 8.8.8.8
nameserver 1.1.1.1

PROBLEM: Proxmox auto-generates this → resets after restart!

Permanent fix (on Proxmox host):

nano /etc/pve/lxc/100.conf

Change: 

# OLD:
nameserver: 192.168.0.102

# NEW:
nameserver: 8.8.8.8

LXC restart: 

pct stop 100
pct start 100
pct enter 100
cat /etc/resolv.conf
ping google.com  # Works! ✅

📊 Final State

Disk usage:

Disk      Size    Used    Free   Use%  Purpose
────────────────────────────────────────────────
disk1     5.5TB   7.6GB   5.2TB   1%   Data (MergerFS)
disk2     5.5TB   11GB    5.2TB   1%   Parity
disk3     916GB   713GB   158GB   82%  Data (MergerFS)
disk4     1.8TB   922GB   818GB   53%  Data (MergerFS)
────────────────────────────────────────────────
MergerFS  8.1TB   1.7TB   6.1TB   21%  Unified pool

Media structure:

/mnt/storage/media/
├── movies/
│   ├── hun/  (654 movies - disk3+disk4 merged)
│   └── eng/  (12 movies - disk3+disk4 merged)
├── tv/
│   ├── hun/  (28+ series - disk1+disk3+disk4)
│   └── eng/  (6 series - disk3+disk4)
├── downloads/
├── books/
└── music/

Docker compose volume mappings (updated):

volumes:
  - /srv/docker-data/jellyfin:/config
  - /mnt/storage/media/movies:/media
  - /mnt/storage/media/tv:/media2:ro

SnapRAID protection:

  • 3 data disks protected (disk1, disk3, disk4)
  • Parity: disk2 (5.5TB)
  • Content files: saved in 4 locations
  • Sync status: OK, 1.7TB protected

⚠️ Important Notes

Boot order:

  1. MergerFS mount (/mnt/storage)
  2. LXC start (this is why a restart was needed!)
  3. Docker containers start

DNS strategy:

  • At this point: Google DNS (8.8.8.8) - temporary fix
  • Later: Start AdGuard Home on dedicated LXC 102 (192.168.0.111) ✅ Done - see Doc 05
  • Then: Modify LXC config: nameserver: 192.168.0.111 ✅ Done - see Doc 05

Scrutiny problem:

❌ Does not work inside LXC (device access restricted) ✅ Solution: Run Scrutiny on the Proxmox host

🚀 Next Steps

Urgent:

  1. ✅ Run Jellyfin library scan
  2. ⏳ Fix paths for other Docker apps (Radarr, Sonarr, etc.)
  3. ⏳ Start AdGuard Home on the new server

Later:

  • SnapRAID automation (cronjob)
  • Monitoring setup (Grafana + Prometheus)
  • Finalize backup strategy
  • Fine-tune network segmentation (arr_stack vs utils)

📝 Lessons Learned

What worked well:

✅ MergerFS flexibility - easy to add new disks
✅ SnapRAID scalability - smoothly scaled from 1 to 3 disks
✅ ext4 filesystem - cross-platform compatibility
✅ Docker compose - simple path modification

Problems and solutions:

❌ Empty folders in MergerFS → Deletion required
❌ LXC mount timing → Restart required
❌ DNS Proxmox auto-config → LXC config modification needed
❌ Docker caches old mounts → Compose down/up required

Session end: 2025-12-23 01:00
Document created by: Claude (Anthropic)