Quick Summary (TL;DR):
To back up Proxmox virtual machines to a NASAliExpress price, create a dedicated NFS or SMB/CIFS share on the NASAliExpress price, add it under Datacenter → Storage in Proxmox VE with VZDump backup file content enabled, run a manual snapshot-mode backup, and then create a scheduled job under Datacenter → Backup. Set a deliberate retention policy, monitor failed jobs and, crucially, restore a backup into a separate VM ID on an isolated network to prove recovery works. Ordinary NASAliExpress price-targeted VZDump archives are full VM backups; they are not incremental merely because the NASAliExpress price uses snapshots or deduplication. For frequent backups of several changing VMs, consider a dedicated Proxmox Backup Server (PBS) for incremental, deduplicated backups and built-in verification. Neither method automatically captures guest-mounted external NASAliExpress price data, PCI/USB devices or every application-specific consistency requirement. A second independent/off-site copy is still necessary.
Choose the Backup Architecture Before You Start
There are two legitimate designs, and they should not be confused. The simple design writes a self-contained VZDump archive directly to a NAS share. The more advanced design sends VM backups to Proxmox Backup Server, which maintains a chunk-based datastore. Both can be selected as storage targets in the Proxmox web interface, but they have different recovery, retention and space-management characteristics. This guide implements direct NAS backup first, then explains when PBS is worth adding.
| Design | What Proxmox writes | Advantages | Trade-offs |
|---|---|---|---|
| NFS share on NAS | Compressed full VZDump archive per backup | Simple; broadly compatible with Linux-oriented NAS; easy file-based export | NAS NFS permissions and export path must be correct; every run produces another full archive |
| SMB/CIFS share on NAS | Compressed full VZDump archive per backup | Convenient on Synology, QNAP, UGREEN and mixed Windows/NAS networks | Credentials, SMB permissions and reconnect behaviour need attention |
| Proxmox Backup Server | Incremental, deduplicated backup data in a PBS datastore | Less repeated transfer, verification jobs, granular restore options | Additional service to maintain; datastore design and off-site protection still matter |
A NAS snapshot of a live VM disk is not equivalent to a Proxmox backup archive. Nor is a Proxmox VM snapshot a substitute for a backup stored in another failure domain. If the mini PC dies or its local SSD fails, you need the guest configuration and disks somewhere recoverable.
What a Proxmox VM Backup Includes — and What It Does Not
A normal VZDump VM backup captures the VM configuration and eligible virtual disks, so you can reconstruct the guest on another Proxmox host. The exact included disks depend on the VM hardware configuration and each disk’s Backup setting. Review VM → Hardware and confirm that important virtual disks are included; a disk deliberately excluded from backup will not magically appear during restore.
| Component | Usually included? | Recovery implication |
|---|---|---|
| VM configuration, virtual CPU/RAM settings and ordinary virtual disks | Yes, for included backup-enabled disks | Restoring the archive recreates the core VM configuration and data |
| Additional virtual disk with Backup disabled | No | Enable its backup flag or back it up separately |
| Data on SMB/NFS shares mounted inside the guest | No, not simply because the guest can access it | Back up the external share independently using the NAS or application tools |
| USB passthrough device, PCI GPU or physical HBA | Device itself: no | Reattach hardware and validate mappings on the destination host |
| Host bridges, VLANs, storage definitions and host certificates | Not as a complete bare-metal host image | Document or separately back up host configuration |
| Database transactional consistency | Not guaranteed by a VM image alone | Use guest quiescing plus database-native backup or application-specific procedure |
This distinction is especially important for Home Assistant VMs, Plex servers, Docker VMs and database hosts. A Proxmox archive can restore the guest OS while a bind-mounted media library, network-mounted documents or external database remains outside the archive. For architecture choices, see our mini PC plus NAS versus NAS-only server guide.
Prerequisites: NAS, Network and Capacity
- Proxmox host: a working Proxmox VE installation with at least one VM and administrator access. The steps match the current Proxmox VE 9.x interface, although labels can vary slightly by release.
- NAS: a powered-on device with enough free capacity, a dedicated shared folder and NFS or SMB service enabled. You can use Synology DSM, QNAP QTS/QuTS hero, UGREEN UGOS Pro, TrueNAS or another NAS that exposes a standard share.
- Network: preferably wired Ethernet and a stable NAS address or reliable local DNS name. Do not rely on a guest VM to provide the DNS service required to locate its own backup destination.
- Permissions: a dedicated NAS share/account or NFS export that the Proxmox host can write to. Do not grant the whole home network broad write access.
- Recovery capacity: enough NAS free space for multiple backup generations, plus a separate target SSD/storage pool on which a restored VM can actually run.
- Protection: a UPS for NAS and host where practical, a tested notification path, and a second copy outside the same NAS.
Keep VM disks on the normal Proxmox SSD/NVMe storage; the NAS share in this tutorial is the backup destination, not necessarily the primary datastore for running VM disks. Backing up a VM that already runs from the same NAS into another folder on that NAS provides little protection against NAS failure. Choose a genuinely separate failure domain when possible.
Step 1: Create a Dedicated NAS Backup Share
Option A: NFS export
Create a shared folder such as proxmox-backups on the NAS, enable its NFS service and grant the Proxmox host IP address read/write access. The export path is NAS-specific: Synology paths often look like /volume1/proxmox-backups, while a TrueNAS export might be /mnt/tank/proxmox-backups. Do not copy an example path without checking the NAS export list. Avoid permitting all clients or disabling root mapping without understanding the security implications. Test permissions with the NAS vendor’s recommended NFS export settings.
On the Proxmox host, discover exports with the storage manager before adding anything:
pvesm scan nfs 192.168.1.20Replace the example address with the NAS address. If the intended export is absent, troubleshoot NAS NFS service, allowed clients and firewall before proceeding. NFS version and root-squash behaviour depend on the NAS; verify that Proxmox can create backup files rather than assuming a successful mount implies write permission.
Option B: SMB/CIFS share
Create a share called proxmox-backups and a dedicated NAS user, for example pve-backup. Give that user the minimum permissions needed to write and read backup archives. Prefer SMB 3 rather than enabling obsolete SMB1. For credentials, use the Proxmox web interface, which stores the secret in its protected storage configuration, rather than pasting passwords into shell history. If the NAS supports share snapshots, configure a separate snapshot policy after testing restore; snapshots are an additional layer, not an off-site backup.
| NAS setup question | NFS | SMB/CIFS |
|---|---|---|
| What to enter in Proxmox | Server IP and actual export path | Server IP, share name, username/password |
| Common failure | Root mapping or host IP not allowed | Account lacks share/file-system permissions |
| Security control | Restrict export to the host or dedicated VLAN | Dedicated account; SMB3; no guest write access |
| Best choice | Often simplest for Linux-oriented NAS exports | Often easiest with existing NAS user/share administration |
Step 2: Add the NAS to Proxmox Storage
- Open the Proxmox web interface and select Datacenter → Storage → Add.
- Choose NFS or SMB/CIFS, matching the share you prepared. Do not select Proxmox Backup Server unless a real PBS instance and datastore already exist.
- Give the storage a distinctive ID, for example
nas-backups. Enter the NAS address and select the real export/share. For SMB, supply the dedicated NAS account. - Under Content, enable VZDump backup file (the internal storage content type is
backup). Disable VM disk images/ISO content if this storage is for backups only. - Confirm the storage is enabled for the node that runs the VMs and save. In a cluster, make sure every selected node can reach the same share and has appropriate access.
- Select the new storage and check its Content and Summary views. It should show available capacity and later display the backup archives.
For an NFS export, the equivalent command is useful for reproducibility. This is an example, not a command to run unchanged; the export must exist and be writable:
pvesm add nfs nas-backups \
--server 192.168.1.20 \
--export /volume1/proxmox-backups \
--content backupInspect the configured storage and verify connectivity:
pvesm status --storage nas-backups
pvesm list nas-backups --content backupThe second command may return no entries before the first backup. That is normal. A storage that is offline, read-only or reporting zero free space is not ready for a scheduled job. If you already mounted a NAS share manually in /etc/fstab, do not accidentally create two independent Proxmox storage definitions pointing to the same directory.
Step 3: Run One Manual VM Backup First
Select a non-critical VM and open VM → Backup → Backup now. Choose nas-backups, Snapshot mode and ZSTD compression if offered. Start the job and read the task log until it reports success. Check the NAS share independently to ensure the archive is really present. On ordinary file-based NAS storage, the archive is normally a vzdump-qemu-...vma.zst file, together with related logs/notes where applicable.
For a VM whose ID is 101, the CLI equivalent is:
vzdump 101 --storage nas-backups --mode snapshot --compress zstdThis example backs up one VM. It does not create a recurring schedule. Verify that nas-backups is your intended storage ID before using it, and avoid running simultaneous heavy backups until you understand how the host SSD, NAS disks and network behave under load.
| Backup mode | How it works | When to use | Main caveat |
|---|---|---|---|
| Snapshot | Backs up a running VM while coordinating a point-in-time image | Default choice for most always-on VMs | Typically crash-consistent unless guest/application quiescing is correctly configured |
| Stop | Stops the guest, backs it up, then starts it again | Critical guests where planned downtime is acceptable | Creates downtime; check shutdown behaviour and recovery |
| Suspend | Uses a pause/suspend-style process during part of the backup workflow | Special cases where the documented behaviour is required | Can interrupt service and is generally less attractive than snapshot for KVM guests |
Important: “snapshot mode” describes how Proxmox captures the VM while it runs. It does not mean the NAS needs snapshot support, and it does not mean the destination archive is an incremental backup. Standard VZDump backups to a NAS share are full archives each time. The exact implementation differs between KVM VMs and LXC containers; this guide is primarily about KVM VMs.
Make Running-VM Backups More Consistent
A backup can complete successfully while a busy database has transactions in flight. Install and enable QEMU Guest Agent inside supported guests, and enable the agent in VM → Options. The agent can help coordinate filesystem freeze/thaw around backup operations when the relevant Proxmox and guest settings are active. On Windows, check the guest-agent/VSS integration and the application’s VSS writer status; on Linux, confirm filesystem freeze support and any application-specific hooks. Merely installing an agent does not guarantee that PostgreSQL, MariaDB, Active Directory or another complex service has produced an application-consistent checkpoint.
- For a small stateless Linux VM, a successful snapshot-mode backup and periodic restore test may be sufficient.
- For Home Assistant, combine VM backups with Home Assistant’s own backups where configuration portability matters.
- For PostgreSQL/MariaDB and other transactional services, maintain native logical/physical database backups or a documented quiesce procedure as appropriate.
- For Windows workloads, validate VSS writers and run a recovery test of the actual application, not just a Windows boot screen.
- For guests with passthrough or external storage, record the extra devices, mount points and network dependencies required after restore.
Step 4: Schedule Backups and Set Retention
Open Datacenter → Backup → Add. Select the node, storage nas-backups, the relevant VM IDs (or an explicit selection mode), snapshot mode and compression. Schedule backups outside busy periods, for example every night after local database maintenance or media indexing. Configure notifications for failures and periodically check the task history. A schedule without alerting can silently fail for weeks after a NAS password change or a full volume.
| VM type | Illustrative frequency | Illustrative retention | Reason |
|---|---|---|---|
| Home Assistant / infrastructure VM | Daily; additional app-native backups after important changes | Last 3 + daily 7 + weekly 4 | Rapid rollback after updates plus older recovery points |
| Small database/application VM | Daily or more frequently according to RPO | Daily 7 + weekly 4 + monthly 3 | Keep application-native transaction-aware backups between VM images |
| Occasional development/test VM | Weekly and before major changes | Last 2 + weekly 4 | Avoid filling the NAS with replaceable environments |
| Irreplaceable production-like VM | Frequency based on measured RPO and restore tests | Documented policy with off-site copy | Do not choose retention merely to match available disk space |
The terms keep-last, keep-daily, keep-weekly and keep-monthly are retention buckets, not a promise of a fixed total number of archives. Proxmox retains qualifying restore points under the combined rules. For example, a policy may be expressed as keep-last=3,keep-daily=7,keep-weekly=4,keep-monthly=3. Review the actual retained list after a few weeks and ensure the policy is applied to the intended backup storage/job. Be careful with storage-level versus job-level pruning settings; confirm which policy is in effect in your version and avoid deleting the only known-good restore point.
The appropriate schedule follows two requirements: RPO (maximum tolerable data loss) and RTO (maximum acceptable restoration time). A daily backup can lose almost a day of changes even if it is perfectly healthy. A 2TB VM may take much longer to restore than a 20GB Home Assistant VM, especially over 1GbE.
How Much NAS Capacity Do You Actually Need?
Size the NAS from measured archive sizes and retention, not the total advertised capacity of thin-provisioned virtual disks. Compression depends on guest data: text and empty blocks can shrink substantially, while encrypted, compressed or media data may shrink very little. Make one representative full backup, record its size, and measure how it changes after normal workload growth.
| Example assumption | Value | Meaning |
|---|---|---|
| Three VM backup archives per daily run | 55GiB + 25GiB + 20GiB = 100GiB | Illustrative compressed archive sizes, not benchmark results |
| Fourteen retained daily-equivalent full sets | 14 × 100GiB = 1,400GiB | About 1.37TiB before NAS filesystem overhead |
| Allow 30% additional capacity | 1,400GiB × 1.30 = 1,820GiB | About 1.78TiB for these assumptions |
| Practical purchase target | At least ~2TiB usable; 4TiB+ offers more growth | Usable after RAID/mirror/parity and other NAS data |
| Independent second copy | Additional storage elsewhere | Not counted in the primary NAS capacity above |
These are deliberately transparent assumptions. Retention buckets may preserve more or fewer archives than “14”, and NAS snapshots can consume additional capacity. For instance, a two-bay NAS with two 4TB drives in a mirror offers roughly 4TB decimal raw usable space before formatting and system overhead, not 8TB. A RAID rebuild or drive replacement does not increase backup history. Our DIY NAS versus prebuilt NAS cost guide explains why disk, enclosure, UPS and replacement costs must be compared together.
Transfer time also matters. Moving 100GiB at a sustained 100MiB/s takes about 17 minutes before backup overhead, compression and NAS disk limits. At 250MiB/s it is roughly seven minutes. These are arithmetic estimates, not measured Proxmox benchmarks: a small NAS with slow disks or an overloaded host may never reach either figure. For network upgrades, see our 2.5GbE switch comparison.
Step 5: Restore a VM From the NAS — Safely
Do not test a restore by overwriting your only working VM. A meaningful test uses a spare VM ID and an isolated bridge/VLAN or a disconnected virtual NIC, so that the restored guest cannot compete with production for the same IP address, hostname, DHCP leases, database endpoints or automation devices.
- Confirm that the NAS storage is online and that the archive appears under Storage → Backups (or the relevant storage Content view). Record the source VM ID and backup timestamp.
- Choose a known-good archive and select Restore. Set a new unused VM ID, for example
201for an original VM101. - Choose a destination VM disk storage pool with sufficient free capacity, such as
local-lvm. Do not accidentally restore large VM disks onto a tiny boot partition. - Before first boot, inspect the restored VM hardware. Disconnect the NIC or attach it to an isolated test bridge; verify MAC address uniqueness and any static IP configuration.
- Start the restored VM. Confirm that the guest boots, services start, files are present and application-specific checks pass. Check that databases, authentication, Docker volumes and relevant network shares work as expected in the test environment.
- Record restore start/end time, actual recovery point, missing dependencies and remediation. Remove the test VM when finished, without touching the original.
A command-line restore is also possible. This example assumes you have confirmed the exact archive filename under the Proxmox-managed mount path. Do not use the placeholder filename literally.
pvesm list nas-backups --content backup
# Replace the archive filename below with a real listed backup:
qmrestore /mnt/pve/nas-backups/dump/vzdump-qemu-101-YYYY_MM_DD-HH_MM_SS.vma.zst 201 --storage local-lvmOn some setups, a file-based backup is presented through a storage volume identifier instead of a path you have typed manually. The GUI is often safer for first-time recovery. Check the qmrestore options for your installed release if you need to regenerate guest MAC addresses or change restore defaults. Restoring a second copy with the original network identity can disrupt a live LAN even when the VM ID is different.
What If the Proxmox Host SSD Fails Completely?
This is the recovery test that matters more than an archive existing on the NAS. You need a replacement host, Proxmox installation media, NAS credentials and the location of your backup storage. The backup archive contains VM data/configuration, but it does not rebuild every aspect of the original Proxmox host. Document bridge names, VLAN IDs, static management IPs, DNS, storage definitions, cluster membership, passthrough device mapping and any custom firewall rules.
- Install a supported Proxmox VE release on replacement hardware and apply appropriate security updates.
- Recreate the management network, VLAN-aware bridges and any storage needed for VM disks.
- Re-add the NAS backup storage with its original share details and verify that backups are visible.
- Restore the most important infrastructure VM first, onto storage with adequate capacity. Reconfigure networking, host-specific device passthrough and guest dependencies.
- Bring back remaining VMs in a deliberate order; validate DNS, authentication, MQTT, databases, Home Assistant and application endpoints before exposing services.
- After recovery, take a fresh backup, investigate the original failure and update the written recovery runbook.
If you use PCI passthrough for a GPU or USB radios for Zigbee/Thread, the restored guest may boot without the hardware until you reattach the correct device on the new host. Avoid hard-coding fragile USB bus numbers when a stable device mapping is available. Back up external NAS shares separately, because restoring the VM alone does not recreate them.
When to Use Proxmox Backup Server Instead
For a handful of small VMs backed up once a day, full VZDump archives on a NAS are often entirely reasonable. When you have many VMs, frequent restore points, slow uplinks or limited storage, PBS becomes more attractive. It uses chunk-based deduplication and incremental transfers, offers scheduled verification and supports convenient restore workflows. The exact space savings depend on changed blocks and how similar your backups are; do not assume a fixed “90% saving”.
| Decision point | Direct NAS VZDump | Proxmox Backup Server |
|---|---|---|
| Setup effort | NAS share and Proxmox storage entry | Separate PBS service, datastore and credentials |
| Backup method | Full archive per run | Incremental transfer with deduplicated chunks |
| Retention | VZDump archive pruning | PBS prune jobs plus garbage collection |
| Integrity verification | Check logs/archive and perform restores | Scheduled datastore verification plus restore tests |
| Granular file recovery | Often requires mounting/restoring archive or guest | Integrated file-restore workflows for supported guests |
| Best fit | Small lab and straightforward disaster recovery | More VMs, longer history, frequent backups and bandwidth efficiency |
Do not treat an arbitrary SMB share as though it were a native PBS datastore. PBS datastores have specific filesystem and metadata requirements; a dedicated local filesystem (for example ext4, XFS or ZFS on appropriate disks) is the straightforward design. If you want to place PBS storage on networked hardware, validate the exact architecture, filesystem semantics, latency, failure behaviour and supportability rather than simply pointing PBS at a CIFS mount. The official PBS datastore documentation explains its layout and maintenance requirements. PBS pruning and garbage collection are separate: deleting backup metadata does not immediately reclaim all deduplicated chunks.
The official Proxmox Backup Server integration guide shows how to add a PBS datastore to Proxmox VE. If PBS is a VM on the same host that it backs up, plan for the circular recovery dependency: the backup server itself may be unavailable precisely when the host fails. A dedicated PBS machine or carefully documented independent recovery path is more resilient.
Security and a Real 3-2-1 Backup Plan
A NAS connected to Proxmox all day is a useful backup destination, but it remains online and reachable. A compromised hypervisor administrator or NAS account could delete or encrypt the archives. NAS RAID protects against some disk failures; it does not protect against accidental deletion, malicious changes, theft, fire or a mistaken retention policy.
- Use a dedicated backup share and limited NAS credentials; do not expose NFS/SMB backup shares to the public internet.
- Keep the NAS and Proxmox on a trusted wired network or restricted management VLAN; patch both systems and disable obsolete SMB protocols.
- Enable NAS snapshots or immutable/write-protected retention where genuinely supported, and ensure backup-writing credentials cannot simply erase every protected generation.
- Keep a separate offline or off-site copy of important restore points; test restoring from that copy, not only from the primary NAS.
- Protect the NAS and networking equipment with a UPS and verify shutdown behaviour, particularly during a long backup.
- Send backup-failure notifications somewhere you will actually read them. Review backup age, free capacity and restore success at least monthly.
For a broader storage protection strategy, our NAS and home-server hardware comparison explains the value of separating compute from storage. If the NAS itself holds the only copy of its configuration or shares, add a separate NAS-level backup as well.
Common Proxmox-to-NAS Backup Problems
| Symptom | Likely cause | What to check first |
|---|---|---|
| NAS storage shows offline | DNS failure, NAS reboot, network/firewall, NFS/SMB service down | Ping NAS by IP from host; check pvesm status and NAS service logs |
| Permission denied on backup | NFS export root mapping or SMB account/share ACL | NAS host-IP allowlist, ownership, SMB user permissions and actual export path |
| No VZDump backup option | Storage Content does not allow backup files | Edit Datacenter → Storage and enable backup content |
| Backups suddenly fail after working | NAS volume full, quota, password change or snapshots filling pool | Check real free space, NAS alerts, credentials and Proxmox task log |
| Backup takes much longer than expected | HDD contention, slow network, compression CPU limit or other jobs | Check source disk I/O, NAS throughput, network link speed and concurrent tasks |
| Backup succeeded but application data missing | Disk excluded, external NAS mount or app-level data outside guest disks | Inspect VM disk backup flags and app-native backups |
| Restored VM boots but service unavailable | Duplicate IP/MAC, wrong bridge, missing passthrough, changed network | Use isolated NIC, inspect guest logs and rebuild device mapping |
| Restore fails for lack of space | Target VM storage is smaller than restored virtual disks | Check target datastore capacity and guest disk sizes, not only archive size |
| PBS space does not shrink after pruning | Chunks still referenced or garbage collection not run | Check PBS prune/GC schedules and active references |
The Proxmox task log is more useful than guessing from a generic “backup failed” banner. Read the first relevant error and confirm whether the failure happened while contacting the NAS, reading a guest disk, writing the archive, pruning old files or finalising the backup. Each has a different fix.
Full Ownership Cost: What to Budget For
Proxmox VE’s integrated backup functionality does not require a separate paid licence, but reliable backup hardware does have a cost. Compare usable storage after redundancy and retention rather than buying the cheapest advertised enclosure. The figures below are a purchasing checklist, not claimed live market prices; regional pricing, drive availability and warranty terms change.
| Cost item | What changes the cost | Avoid this mistake |
|---|---|---|
| NAS chassis | Number of bays, CPU/RAM, NIC speed, warranty | Buying a large application NAS when only a simple backup target is needed |
| Backup HDDs | Usable capacity, CMR specification, redundancy, replacement stock | Confusing raw two-disk capacity with mirrored usable capacity |
| Network | Existing 1GbE versus 2.5GbE switches/NICs | Upgrading networking without measuring the backup window |
| UPS | Actual watts, runtime and shutdown integration | Assuming VA rating equals usable watts or battery runtime |
| Independent copy | USB disk rotation, second NAS or cloud storage | Keeping every backup generation on one powered-on box |
| Operating costs | NAS and switch idle watts, electricity rate, drive replacement | Ignoring 24/7 electricity and future capacity growth |
| Recovery effort | Time to reinstall host and test guest services | Equating “backup complete” with an acceptable RTO |
An illustrative always-on backup NAS averaging 30W uses 0.03kW × 24 × 365 = 262.8kWh/year. At an assumed electricity rate of €0.25/kWh, that is €65.70/year, or about €328.50 over five years before hardware, replacement disks or UPS losses. This is an example calculation, not a measured NAS draw or a claim about your electricity tariff. If your NAS is already running for other services, calculate only the incremental power and storage cost attributable to Proxmox backups.
A Practical Restore-Test Checklist
- At least one scheduled backup completes and the NAS contains a real archive.
- The intended VM disks are included; excluded disks and external shares are documented.
- The retention policy keeps multiple useful recovery points and leaves capacity headroom.
- Failure notifications arrive when a job is deliberately cancelled or a test condition is triggered (without risking real data).
- A backup is restored under a different VM ID to a test datastore.
- The restored VM boots with an isolated network identity and passes application checks.
- The measured restore time and actual recovery point meet your stated RTO/RPO.
- You can recover even if the original Proxmox boot SSD is unavailable.
- An offline or off-site copy is readable independently of the primary NAS.
Verdict: Start with a dedicated NFS/SMB NAS share and a tested nightly VZDump job if you have a small home lab. Upgrade to PBS when backup frequency, archive growth, verification needs or restore flexibility justify another service. In either case, the most valuable step is a controlled restore test: it exposes missing guest disks, host configuration and network dependencies before a real failure does.
Related esp32.co.uk/ Guides
- Mini PC Plus NAS vs a NAS-Only Home Server — understand separate compute/storage failure domains.
- UGREEN NASync vs Synology for Home Assistant — compare NAS software and hardware trade-offs.
- Best 2.5GbE Switches for a Smart Home and Home Lab — improve backup-network throughput where it matters.
- DIY NAS vs Prebuilt NAS: Complete Build and Running Costs — include disks, UPS and operating costs.
- How Much RAM Does a NAS Need for Docker, Plex and VMs? — avoid overspending on the backup appliance.
Official Documentation and Further Reading
- Proxmox VE — Features and integrated backup/restore — full VZDump archives versus incremental PBS integration.
- Proxmox VE — Storage manager documentation — NFS/CIFS storage and the backup content type.
- Proxmox VE — Backup and Restore administration guide — modes, scheduling and retention.
- Proxmox VE — VZDump command reference — command options for manual backups.
- Proxmox VE — VM restore command reference —
qmrestoresyntax and options. - Proxmox Backup Server — VE integration — add PBS as storage.
- Proxmox Backup Server — Datastore requirements — filesystem and datastore considerations.
- Proxmox Backup Server — Maintenance and verification — prune, garbage collection and verify jobs.