Space‑Optimization Plan for pve-dev

| Category | Current Status (from `df -h`) | Optimization Ideas |
|———-|——————————|——————–|
| **Root** (`/dev/mapper/pve-root`, 211 GB) | 84 % used (169 GB) | • Clean APT cache: apt-get clean
• Remove unused packages: deborphan --clean or apt autoremove
• Find large logs (`/var/log/*.log`) – rotate or truncate.
• Delete old snapshots of containers/VMs (e.g., via `pct delete-snapshot`, `qm deletesnapshot`). |
| **LXC Containers** | All running; many may not be needed. | • Audit each container’s disk usage: pct du <vmid>
• Stop and remove unused containers (`pct stop <vmid>; pct destroy <vmid>`).
• Compress or delete unused local templates in `/var/lib/vz/template/`. |
| **VMs** | Many VMs listed; some are stopped. | • Delete stopped VMs that are no longer required (`qm destroy <vmid>`).
• For large disk images, consider moving to thin‑provisioned disks or using qemu-img shrink.
• Consolidate snapshots: delete older snapshots that aren’t needed. |
| **NFS mounts** | Plenty of free space – no immediate issue. | • Ensure that backups are still required; if some backup volumes can be purged, move to cheaper storage. |
| **Logs & Temp Files** | `/tmp` and `/var/tmp` are almost empty, but check `/var/log`. | • Configure logrotate for any custom logs.
• Remove old kernel images (`dpkg -l ‘linux-image-*’` – keep only latest 2). |
| **Proxmox Backup Server (PBS)** | Not directly on this node but may hold backups. | • Check PBS usage; delete obsolete backup archives via the PBS web UI or CLI. |
| **Package & Container Images** | `/var/lib/vz/template/` can grow large. | • Remove unused templates and ISO files. |

### Quick Commands to Run (in order)

# 1. Clean APT cache & auto‑remove packages
apt-get clean
apt-get autoremove --purge

# 2. Find largest directories in root
sudo du -hX dev / | sort -rh | head -n 20

# 3. Remove old kernel images (keep latest 2)
dpkg -l 'linux-image-*' | awk '/^ii/ {print $2}' | grep -v "$(uname -r)" > kernels.txt
apt-get purge $(awk 'NR>1{print $0}' kernels.txt)

# 4. List LXC disk usage
for vmid in $(pct list --output-format csv | cut -d',' -f1); do echo "$vmid:"; pct du $vmid; done

# 5. Delete unused containers (example)
pct stop 600
pct destroy 600

# 6. List VM snapshots and delete older ones
for vmid in $(qm list --output-format csv | cut -d',' -f1); do qm listsnapshot $vmid; done
# Then: qm delsnapshot <vmid> <snapshot_name>

# 7. Remove old backups from PBS (via API or UI)

### What to Verify After Cleanup
| Check | Command / Tool |
|——-|—————-|
| Disk usage refreshed | df -h |
| LXC list after deletions | pct list |
| VM list after deletions | qm list |
| APT package status | dpkg --list | grep '^ii' | wc -l |

### Expected Impact
– **Root filesystem**: 10–20 GB freed (depends on number of old kernels, logs, and unused packages).
– **Containers/VMs**: Potentially >50 GB if many are removed.
– **Overall**: Bring root usage below ~70 % and free up space for future workloads.

—

Feel free to adapt the commands to your environment or let me know if you need a more detailed audit script!

By General