• 0 Posts
  • 10 Comments
Joined 11 months ago
cake
Cake day: August 16th, 2023

help-circle




  • The cost of something isn’t always in the form of money. In many cases with Foss there are comprises in either simplicity, stability, documentation, or compatabiliry.

    For instance I can boot my machine into a live garuda instance and it runs great, but as soon as I install it, it runs like trash. I spend something like 3 hours fiddling trying to get it going then wipe and try to install smaugos and it wont even boot. I install debian and it works okay but sluggish. Popos works fine. 2 days of fiddling around and I find something that works. Windows may cost more than just money, but it worked out of the box and I didn’t have to fiddle or try a bunch of different distros. We can go down that rabbit hole, but let’s look at other things.

    Foss often has volunteer support that can be hit or miss and often requires more advanced knowledge of the os or software. There’s also often toxicity like people shaming for not knowing everything about the application or os. Commercial support is often dedicated and may even remote into your computer. I’m not saying Foss can’t do that, but I’ve never heard of it for free.

    FOSS doesn’t work nearly as easily or reliably as commercial software a lot of the time. Nextcloud is a good example. There are a million ways to install it, but now you need to learn docker, or how to setup a web server and even then maybe the docker image is buggy or straight up doesn’t work. The different Linux distros is another example.

    Then there’s the learning curve. Even if FOSS has 1:1 parity in functionality, it often comes at the cost of learning a LOT about a new application, or the functionality is different or harder to use compared to a commercial alternative.

    Don’t get me wrong I live foss. I self host, I’m slowly getting rid of windows and degoogling. But there is cost to do all of this, even if its not monetary. Plus not everyone has the time, patience, or interest in it.



  • I started with truenas core, then moved to truenas scale. I tried a couple others but ultimately truenas had an easier and cleaner ui and i wanted an easy way to use nextcloud without having to do too much work.

    Turns out nextcloud doesn’t seem to work right now, probably a user error in the container deployment, so I’ve not done that again. Most of the containers I’ve tried using I end up just building vms for because it’s more flexible. Right now I have 2 720xd one is truenas storage, the other is proxmox. They’re both on 10g network to a switch so using truenas to store data for the vms on my proxmox isn’t a big deal at all.

    In any case like I said I don’t really use the truenas box for much other than storage which is a shame, there’s a boat load of memory and like 32 cores. Currently I backup to USB drives. Not great, but I also don’t want to burn my money on cloud storage or hefty external raid enclosures. Tape would be cool, but again I’m a cheap boi.

    When it comes down to it, this is what is recommend. Write down a list of what you’re requirements are and what you’d like to see. Compare the filer oses and pick ones that meet the requirements and what you like. Then just install them and see what the look and feel is.

    Don’t forget backups, people will preach gospel about needing 1x2x3 or some sort of other potentially expensive backup solution. If this is a home lab, do what fits your budget, skill, comfort levels. You can always improve from there. External drives work fine for me, will they both fail at some point sure, but nothings perfect and more important data is backed up to encrypted blobs in free cloud storage.

    Also remember to take your time. It’s easy to Leroy Jenkins some shit and just go in guns a blazing, but if you take your time and read and make sure you understand the important stuff before you implement, you’ll save a ton of time. Unlike me who had to blow out my zfs impingement once after 5 tb were uploaded and kept screwing up my backups. Glad I didn’t loose data, but easily could have happened.