Not sure I’d use either, but may I suggest both, with a clear caveat that one or both may disappear or change ? Throw stuff at the wall, see what sticks…

You can cycle the smaller drives to cold backup, that’s not a waste. You do have backups, which RAID is not, right?

Sure, works fine for inference with tensor parallelism, USB4 / thunderbolt 4/5 is a better (40Gbit+ and already there) bet than ethernet (see distributed-llama). Trash for training / fine tuning, that needs higher inter GPU speed, or better a bigger GPU VRAM.

Fair enough, I got the wrong impression with the post (which I did read) finishing with

It is not impossible to circumvent these issues, for instance by paying for a jmp.chat phone number with monero XMR. In this case you don’t actually have the sim, but rather access it remotely over XMPP. If you do this over Tor very little can be used against you.

and so on, which I found wanting from a privacy perspective at the implied threat level. No phones (or perhaps faraday bags, or aircraft mode, if tested, depending on threat model) is a much more astute take home, hence the spycraft suggestion.

Anyhow, best of luck.

So, your key takeaway seems to be getting an untraceable phone number. From an opsec point of view I see a few problems.

First, this is implicitly aimed at going against state level actors, which is a whole other game than random internet services. With that in mind…

You assume TOR is actually anonymous, but it has been shown that with enough compromised exit nodes that fails. It’s also a NSA project originally, which may or may not be relevant, the code may be good and is open and has had eyes on, but at the least shows they are intimately familiar with it.

You assume acquisition of Monero is uncompromised and untraceable. Perhaps cash at a machine might be pretty good, but a camera could easily invalidate it, or the machine itself be compromised, wouldn’t be hard to imagine a profit motive or false flag driving that.

What’s the security implications of the XMPP protocol ? Just using TOR may not be enough (I don’t know, just asking the questions). What about the other end of the phone call?

One approach, especially for local efforts, is just using old school spycraft, dead drops, one time pads etc.

You asked for feedback.

Until next week, sigh.

Huge kudos to Freetube, I love it. Just the opposite to Alphabet, go on, be evil again.

Seems like data integrity is your highest priority, and you’re doing pretty well, the next step is keeping a copy offsite. It’s the 3-2-1 backup strategy, 3 copies, 2 media (used to mean CDs etc but now think offline drives) 1 offsite (in case of fire, meteor strike etc), so look to that, stash a copy at a friends or something.

In your case I’d look at getting some online storage to fill the offsite role while you’re overseas (paid probably, but a year of 1 or 2 Tb is quite reasonable) leaving you with no pressure on the selfhosting side, just tailscale in, muck around and have fun, and if something breaks, no harm done, data safe.

I’ve done it for what seems like forever and I’d still be worried about leaving a system out of physical control for any extended period of time, at the very least having someone to reboot it if connectivity or power fails will be invaluable, but talking them through a broken update is another thing entirely, and you shouldn’t make that a critical necessity, too much stress.

Thing is, the time for net-zero has passed, did you hear that whooshing sound?

To pull back from the brink, what is needed is net-negative, which ain’t happening without capture (alongside massive reduction in emissions), economics be damned, it’s an existential threat, it’s about survival. Could be as simple as massive reforestation, could be fusion generators pulling CO2 out of the air, will probably be many different things, but learning what works, as soon as possible, is imperative.

IMO, same reason they have their own repo, which eventually feeds into Red Hat enterprise, to have a trustworthy, curated set of safe (ish) software that’s had eyeballs on it. A worthy enough goal, but that said, it applies a lot less to flatpaks. I personally used to remove theirs because I didn’t like having multiple sources, now I’m on Bazzite which ships with flathub.

The old adage is never use v x.0 of anything, which I’d expect to go double for data integrity. Is there any particular reason ZFS gets a pass here (speaking as someone who really wants this feature). TrueNAS isn’t merging it for a couple of months yet, I believe.

Yup (although minutes seems long and depending on usage weekly might be fine). You can also combine it with updates which require going down anyway.

Basically, you want to shut down the database before backing up. Otherwise, your backup might be mid-transaction, i.e. broken. If it’s docker you can just docker-compose down it, backup, and then docker-compose up, or equivalent.

Would watch that show in a heartbeat.

Solid. My backup is a T440p, and behind that a X230, fucking bulletproof.

Been happy with FreshRSS for years now (TTRSS before that). One thing that really improves it is RSS-Bridge which turns a lot of non RSS sites into feeds (and a lot of truncated feeds into full ones). It’s also a list of what hackerly types will put effort into getting a feed from, so self-curating in its way. Enjoy…

No Gods, No Masters

Also, prefer the Stargate approach, but recognise Star Trek is likely more responsible.