Backing up your data is one of the most common pieces of advice you’ll hear on the Internet, and there’s a good reason why many veteran tinkerers keep harping on this point. After all, it’s hard to realize the significance of frequent backups until you end up in a situation where you can’t recover your systems without redundant copies of your precious files.

Although backups are essential for casual users, their utility jumps to the next level in home labs filled with VM-based projects and containerized, self-hosted applications. Even if you’re not as much of a data hoarder as I am, there’s a lot of data you’d want to preserve when you’ve got a home lab.

Virtual guest data

PBS for Proxmox, Veeam for everything else

Although virtual machines and containers are quite disposable, there’s no denying that they’re the backbone of every home lab. When you’ve spent hours painstakingly configuring your virtual guests, you wouldn’t want to lose them permanently to a botched project. I’ve got dev VMs and dozens of self-hosted services that I consider essential for my everyday tasks, and rebuilding from scratch would be a royal pain.

While the amount of data you can save depends on the storage at your disposal, you should at least create backups of your essential virtual guests. For Proxmox users, I recommend going with a PBS instance running on a separate machine, as the first-party tool not only makes your snapshots space-efficient, but also provides quick ways to recover your LXCs and VMs when things go wrong. But if you’re on XCP-ng, Harvester, ESXi, or other home lab platforms, Veeam’s backup agent is compatible with most – either directly or via plug-ins. I’ve used Veeam with XCP-ng and Proxmox in the past, and the free version of the app works pretty well for normal backup operations.

System config files

Of the underlying server distro

When your home lab projects involve extensively modifying the network settings or storage permissions, your chances of ending up with a failed project where even the virtualization platform is rendered unusable become rather high. If you’ve got recent backups of your virtual guests, you can spin them up after reinstalling the underlying distro. Unfortunately, your old storage directories, user groups, network configs, and other essential OS files will get wiped out by the fresh installation.

An easy solution involves backing up the configuration files beforehand. Unlike massive VM and LXC data, most config documents tend to occupy less than a megabyte’s worth of storage space. I tend to use tis24dev’s backup script to create extra copies of my PVE and PBS configs, while Harvester’s reliance on YAML documents makes backing up the virtualization platform fairly simple.

Home lab documentation

Your very own source of truth

Documenting every aspect of your home lab may sound tedious… and well, journaling details of server experiments, VM projects, and self-hosted containers requires some patience. However, your notes are bound to come in handy if you need to track changes or retrace your steps when troubleshooting.

But if you rely on containerized tools for documenting your servers, you might want to back them up as well. For example, I tend to save my project notes inside a self-hosted Trilium PKM, while a NetBox instance holds all records of my home network. Although I’ve deployed both tools on a Raspberry Pi instead of my server nodes, backing them up ensures my notes will remain safe even in the case of catastrophic failures.

Media files

Ebooks, photos, ripped ROMs, movies – you name it

Using self-hosted services to organize media is quite popular in the home lab community – to the point where you’d find multiple applications that cater to certain file formats. You’ve got photo managers, Ebook library builders, local video streaming tools, ROM managers, and music servers, with certain utilities like NVR platforms generating new data practically every day.

But if you’re anything like me, you probably rely on network shares or (at the very least) separate datasets to store the media you manage with the containers. Unlike virtual guests, media files can easily consume thousands of gigabytes worth of storage space, making them even more important to back up. Since I’ve got a NAS hosting all my media files, I just sync the essential datasets with a remote storage server. Speaking of remote syncs…

Backup data

Secondary backups are just as essential

The biggest gotcha moment in my early home lab days was learning that a single copy of data is meaningless. Since unforeseen hardware failure is all it takes to make all your backups irrecoverable, it's always a good idea to create redundant copies of backup files. Better yet, if you’ve got two NAS units, you can turn one of them into a remote, secondary storage server for a 3-2-1 backup setup.

That way, you’ll still be able to access the essential data – including backup files – should something happen to your primary server. Me? I rely on Tailscale for my remote synchronization needs, as it works exceedingly well with PBS, TrueNAS, and every other backup/NAS platform in my computing arsenal.

You don’t need to invest in expensive backup rigs, either

With the ever-increasing prices of storage drives, creating a dedicated backup workflow may seem like an expensive deal. However, you don’t need the most overkill NAS armed with larger-than-life hard drives. An old PC can double as a reliable storage server for your home lab, and you can pick up cheap mini-PCs (or even SBCs) for remote-based secondary backup rigs.