I was reading a homelab discussion about NUTs (Network UPS Tools) that left me scratching my head and wondering "…why? Is complexity for complexities sake part of homelabs? Isn’t this a huge overkill for one machine? Just…use a UPS? "
Then I got to thinking more broadly about homelabbing and I started to wonder if there weren’t maybe (at least) two different schools of thought.
Using home media as example -
- Store the original.
- Detect the client.
- Transcode when required.
- Monitor the GPU.
- Add reverse proxy et al
- Track bandwidth.
- Add user accounts
- Add failover.
- Graph the result.
- Dashboard.
Vs
- Store a common compatible file for your devices.
- Play it directly.
Which school of thought are you and why?
PS: not throwing shade. I get it; for some people, complexity, learning infrastructure, practicing skills for work etc is part of the why. For me, complexity as recreation is suspiciously like work after work.
Perhaps there’s an odd Venn diagram between homelabbing, interest in ownership / useful capability and infrastructure that could make for a fun discussion. For you, is homelabbing a means to an end or is it an end in itself?
EDIT: Feel free to consider “home-lab” and “self hosted” as synonyms for the purpose of this discussion (although I am aware they are somewhat different in scope).
So, you don’t need to run any software at all on a host on a UPS. You don’t need NUT or anything similar. You can just plug a PC into a UPS.
However, traditionally, the purpose of a UPS isn’t “just power a machine through an extended outage”. It’s to provide enough time to cleanly shut down a machine. Save work, and—once upon a time, filesystems that weren’t cleanly unmounted could become corrupt—unmount filesystems.
It can also provide time for something like a backup generator, if you have something like that up. The UPS may only have enough juice to run the PC for a few minutes. If you look at traditional UPS units, they have very limited amounts of time that they can keep a PC running. They aren’t aimed at the “whole house multi-day keep things running” capability that some lithium ion battery setups are these days.
So traditionally, one would have software on a machine that takes some action, like shutting it down cleanly, when the battery gets low.
If what you have a UPS for is to keep a PC running through momentary power interruptions, and you don’t care about, say, remotely monitoring power status, then, sure. NUT isn’t necessary. Same thing if you just want to provide a short amount of buffer time to let a generator that comes on automatically in an outage fire up.
But if you want remote monitoring of power status or you want hosts to shut down cleanly, then it may be useful. For some people, that’s the reason that they have a UPS at all.
Exactly. Which made me wonder why someone would use this for a small machine sitting by their feet, that didn’t serve external anything. Horses for course and all that.
Modern filesystems can corrupt as well with unclean shutdown. They are just less likely to do so. My server stores photos of our life over couple of decades with other nearly irreplaceable data so having the UPS and clean shutdown on power loss gives at least some confidence that the files stay where they are.
Of course I also run backups of the data, but it’s another layer on data safety setup.
It shouldn’t be possible for them to become corrupt on an unclean shutdown, at all. Not at the filesystem level. It isn’t just a question of probability.
It’s possible that there is some bug in the implementation of a filesystem that causes it to violate its guarantees.
There could be, oh…hardware bugs, say, where hardware violates guarantees.
Files can be partially-written. Software might not handle that, though properly-written software designed like a DBMS will be set up so that a partially-written file will not become corrupt at the file level.
Some filesystems permit disabling features that guard against filesystem corruption.
But absent something like that, it should not be possible for one of the filesystems to become corrupt on an unclean shutdown, at all.
Not all modern filesystems are CoW like btrfs. ext4 is merely journaled, which means that there are rare instances where it can be corrupted, but recovery is fast. However in both cases you can still end up with individualy inconsistent files if, say, the power goes out while the CoW b-tree is percolating up and so the file never gets updated or if the ram cache never has a chance to flush.
None of these things mean that the filesystem becomes corrupted; the filesystem will be in a filesystem consistent state when remounted. You can have a write that is partially performed on a file, but that does not cause the filesystem to become corrupt when it does not complete being written. The journal can contain that write, but it can not yet be committed, and won’t yet be visible to a user. But that doesn’t make the filesystem corrupt; the commit itself is atomic.
Well, they might want that save-data-and-clean-shutdown functionality. I mean, I don’t personally care all that much for my use case, but it’s a legit want. Like, say you have a USB flash drive plugged into your machine. Those normally ship formatted with exFAT, a Microsoft filesystem derived from the elderly FAT16 filesystem, and most people use them like that, as pretty much every OS out there can read FAT. FAT can become corrupt in the event of power loss. Or say you’ve been working on something in some application that doesn’t do auto-save and have a day of work accrued in the thing and then you get a power loss. Having a few minutes notice to clean up and shut down is legit important, same way that a laptop will issue low power warnings on a battery. This makes sure that you get that notification.
I think that in general, the clean shutdown functionality isn’t as important these days as it was in the 1990s, as most machines are using journalled filesystems that don’t become corrupt on power loss these days. NTFS (Windows land), HFS+ (MacOS X), or ext4 or btrfs (Linux) for most of their storage. But, I mean, there are going to be people who want it. I don’t think that someone has to be wanting complexity for the sake of complexity for them to have a use case where that applies.
EDIT: Looks like HFS+ actually got replaced with a newer FS a few years back as the default on MacOS.
And even if the filesystem doesn’t become corrupted, individual files can become inconsistent or incomplete. Such as if two files reference each other, but one completes its write to the new version while the other is still in flight when power goes off.
Sure, but that is not filesystem corruption, and I addressed it in my comment.
But in the real world it doesn’t really matter if it’s the whole filesystem or an idividual file which is corrupted. Either way the data is compromised and nitpicking that the filesystem is actually just fine doesn’t make any difference.
The big concern with filesystems becoming corrupt is that all of the data can be lost on a filesystem. Most of the time, this doesn’t show up, but if you lose power at just the wrong time, it can happen. With FAT, HFS, ext2, and similar filesystems that were vulnerable to corruption on power loss, that was a real risk.
So having some kind of power source to make an unclean shutdown unlikely avoids the risk of a really, really bad scenario.
The general shift away from reliance on those filesystems has mitigated the really-bad-case that drove a lot of UPS use.
Databases are probably the most-critical case where applications update a file on disk constantly, and where corruption would be bad. As long as they live on a filesystem that itself can’t become corrupt on power loss, they’re normally structured with a carefully-ordered series of writes and use of write barriers so that they cannot become corrupt.
An individual file can be partially written out. Like, say I’m saving a file in Photoshop, and I have only written half of that file. What winds up on the disk is the first half of of a Photoshop file. That’s an invalid Photoshop file. For some applications, that can matter. Like, say (a) I chose to overwrite an old file and (b) the application doesn’t atomically store the new file (like, write it out, fsync(), and then rename the new file over the old) and © there’s no backup or other system in place.
Or, even aside from partially writing out a file, I can entirely lose work that isn’t saved at all, if I’m not using a software package that makes use of some sort of autosave system. If I haven’t saved all day, then, well, all of that unsaved work is going to go away if there’s a power loss.
That’s a much smaller potential data loss, but, sure…power loss can cause data loss, even if it doesn’t produce that “everything is lost” filesystem corruption.
I’m not asserting that there’s no utility to a UPS; quite the contrary, as my comment was explaining to OP why there is use some people have for having NUT even if he personally doesn’t care much about it for his use case. I’m just pointing out that some of the major reasons for having a UPS+host-side-software combination that existed in, say, the 1990s aren’t really much of a thing in 2026. That was that you really did not want machines using filesystems normally used then to perform an unclean shutdown. Even with a UPS and host-side software like NUT, you can have data loss, as NUT cannot force all software to save unsaved data; there’s no standardized mechanism for forcing applications to save. If the PC is attended, you might get a warning and have a chance to manually save, but that’ll be the best that’ll happen, and if the PC is unattended, well, that save isn’t going to happen. But what NUT and similar software can do is to, when the UPS battery level gets low, trigger a clean shutdown. That ensures that the OS has time to perform a clean shutdown, and on filesystems that have a risk of filesystem corruption if the power is cut—FAT, HFS, ext2, and similar—a UPS with host-side software would avoid that case from arising.
If someone wants to have a UPS in 2026 on filesystems that don’t have those limitations, that’s fine. They may want to have, say, a five minute window to save their work if they are at the PC when power loss occurs. It may help avoid problems that arise from software that cannot cleanly handle power loss—I that Steam can, if power loss occurs while it is updating itself, wind up in a broken state from which it cannot automatically recover when restarted, as I’ve done that myself.
But generally, if you lose power on a system in 2026, you aren’t facing a “I might lose the contents of my hard drive” scenario.
There’s plenty more reasons than filesystem corruption to use clean shutdown and even just plain UPS without any status information to the host. I haven’t seen a full drive corruption due to power loss ever. Various levels of corruption, yes, but not ‘all the data is gone’ scenario due to filesystem being broken. But I have seen multiple hard drives which have died on host shutdown without any warning (no smart errors or such), after power loss they just refused to start again.
Just today we had multiple short (~30 second) power outages and without UPS that would’ve put considerable strain on my hardware. Gladly those are pretty rare and those were well short enough that my UPS kept the system running, without one that would’ve meant that my hypervisor would’ve crashed and that would’ve resulted on several unclean reboots for the VMs.
Also, back to filesystems, with selfhosting, some kind of hypervisor is often in the picture, so the drive activity is generally much higher than with single workstation and corruption on a single file or some part of filesystem is way more likely. Total data loss is pretty unlikely, but a corrupted VM image can be a pretty big deal, regardless if it’s filesystem or single file which is the main issue.
On Linux (and I guess BSD and other *nix variants) there is. SIGTERM is exactly for that situation. Sure, programs can choose to ignore it, but I don’t think I’ve ever seen at least any FOSS software which wouldn’t respect that. Of course the program might crash and cause loss of data, but the mechanism is there.
You’re of course generally correct about modern filesystems being more resilient to power loss, but they’re not immune to problems and also you’ve omitted a large chunk of other reasons why people like to have UPS on their system, with or without monitoring them.
For me personally the hardware is least interesting part of running a homelab. I’d happily run everything on VPS if there was no constraints on money or bandwidth (among other things). But as the limitations are there running things on my hardware is a necessary evil and for that an UPS or two are practically must to have to prolong life span of hardware and prevent data losses. Of course that’s just one box to check on the stack, but when planning for a home server setup for anything I’d say it’s a pretty vital part for the whole setup.