Hi all,
Quite excited for my first post here and for being able to join the club :)
I recently bought a Beelink Mini S12 Pro and put Debian on it. I did use yunohost for over a year and I experimented with basic nginx, fail2ban etc. on a vps but I can’t call myself an expert around these things. For this reason I installed casa-os to ease my way into docker (maybe in the future I will do everything DIY, I do like the idea). I put everything behind tailscale to be able to access the server outside my home.
My idea is to have a media server (radarr and sonarr for movies and tv shows, plus navidrome for my music collection) and some other niceties like audiobookshelf and maybe immich. I still need to learn most things about the arr stack, transcoding and such, so I don’t have any rush.
However, for all these things, I definitely need a larger drive, since the mini pc comes with a 512gb internal ssd.
The S12 also has a 2.5" drive bay (SATA3), but should I go for an internal drive or an external one connected via USB?
Things I am worried about the internal solution is mostly temperatures, everything is so clumped in there and I read some stories about faulty hardware basically heating the sata ssd to death after a while. On the other hand, usb connection is maybe less “stable”? I don’t know. Also, if I go with SATA drive, should I go for ssd or hdd? I don’t have too much budget to invest right now, so cost is somewhat a limit at the moment.
Seeking any advice here and happy to hear more expert people opinion on this, thanks! :)


2.5" is going to limit the storage size a lot. So if you need >4TB, the internal option isn’t going to work. A janky solution would be to route the SATA and power connector outside the mini-pc so that you can fit a larger drive, but that’s going to be pretty ugly, and still limit you to only a single drive.
A single hdd can be fast enough to serve UHD ripped 4k video, but it’s much closer to the limit than an ssd would be, and might not be a great experience if you will be doing anything else with that drive at that time. Having a cache drive or multiple hdds in an array (or even better, both) greatly improves this.
A usb enclosure would let you easily have multiple hdds, but as everyone will say, they are less reliable. My opinion is that while you should never use one as a boot/system drive, they’re fine for bulk storage for home use. Make sure you’re not writing logs or anything like that to it, it should be on demand use only, and you might have to reconnect it occasionally. Anecdotally, I’ve never had issues with usb enclosures, they’ve worked fine for me in the past and I continue to use one for backups, but maybe people with some horror stories would have very different views on this.
You might be mixing up your units here as 4k UHD is typically <90Mb/s while large HDDs typically cap out at 120MB/s, which equates to 960Mb/s (bits vs bytes). You could likely stream 10 4k UHD movies at once from a single HDD before running into bandwidth issues and with the cost of SSDs versus HDDs, it’s almost a no-brainer to go with an HDD.
Thanks for the detailed answer! :) Indeed, I read that I would be limited in storage size if choosing the internal option. That said I am not sure I will be using more than 4tb in the next months, considering also that I don’t have the need for 4k videos at the moment. Would it be a “waste” to just go for an internal ssd to start, and then upgrade to a larger hdd (external) in the future?
Also, sorry just to clarify, when you say
Does downloading media to the drive still counts as a viable thing to do? And I guess the on demand use would be e.g. streaming.
Sorry, on demand is not a good way to state this, it’s just how my weird mind thinks of things. By “on demand”, I mean, like you are actively using it to store something or view something. If you’re not intentionally doing something with it, the drive should be completely idle. That’s more of a target than a requirement, though. It’s a way to keep storage drives tidy and not littered with temporary cache files, or databases used to store runtime state by various services. It’s just a strategy I like to take, to keep bulk storage separated from the applications and services that use it.
Even if a usb drive is intended to be permanently attached, it should still be treated as a temporary component. The reason is so that if something happens and the drive is disconnected, it limits the disruption to the system. You lose your media and documents until it’s reattached, of course, but the computer keeps chugging along happily.
If you use it for writing log files, then its loss can disrupt those services (and also prevent the problem from being reported). Also it’ll be constantly making noise, which can be annoying.
That’s my reasoning, anyway, you might prefer it done differently.