If you plan to expand further, or add another node, I would recommend starting with zfs and see how it performs, because it’s integrated with proxmox and is required for native replication. But you can’t safely convert a zfs mirror to raidz1, you have to take one of the mirrors offline, create a degraded raidz1 with only two disks, copy the data, then wipe the third disk and add it. It’s sketchy and should not be performed with data you can’t afford to lose. But it works, I’ve done it.

Second to that, I’d do traditional software raid, though you’d probably end up doing the same process.

I personally don’t like lvm. I find it frustrating to work with. I’m not sure if you could do the same operations as above.

I’m not sure if you can do data recovery on proxmox virtual disks, but even if the host is completely dead, you can reinstall it and import the VMs and disks. I’ve done that too. But now I run pbs and rclone the backup files to the cloud.

Yes. Browser engines are a hell of a lot more forgiving. And a lot faster to iterate with during development.

I’m not familiar with n8n, but any time you accept user input, it’s dangerous. What happens if a user submits 10,000 emails per second? What if they submit user@example.com'; DROP TABLE emails; --, or whatever the n8n equivalent of SQL injection is? What if they submit ,,,,,,,,,,,,,,,,? What if they submit a blank field? What if they submit completely invalid random binary data? What if they submit a very, very, very long email address?

Debatable. But I have it enabled on my Android phone, it and indeed only shows me the last 24 hours.

If you accidentally dismiss a notification, you can go back in the history to see it. Or if you dismiss a message notification that you want to respond to later. Or if a notification keeps popping up and disappearing and you want to investigate.

Yeah but those aren’t going to be running Linux 7.0. I’d be shocked if they ran anything newer than 2.6, if that new and Linux at all.

The biggest reason is to prevent iot or other untrustworthy devices from reaching the Internet.

That’s a different vector.

Okay so… How do you have it set up and configured? You’ve given us nothing to go on.

Spill your beer on it and get all three in one go!

Anything in the logs when you plug it in?

If it doesn’t, it can still be a paperweight!

I’d still give it a shot. A quick check of benchmarks suggests it’s not that much slower. I don’t know if that extends to ML computation though.

I was playing with ministral-3 3b on a 3060. It loads pretty quick, but response generation is a bit slow. It starts responding nearly instantly once the model is loaded (which is also quick), but for long responses (~5 paragraphs) it may take 15-20 seconds for the whole thing.

Netbird is great and very user-friendly. Well, relatively. I had to bumble my way through the setup the first time and redo it, but certainly a lot easier than headscale/tailscale.

Why do you want to set this up, though?

It probably works.

If you have a very steady hand, you can cut the PCIe slot at the back to allow longer cards to fit, too. Most of them support running at reduced bandwidth (though I would check for x1 specifically, they might only go down to x4).

Tubearchivist maybe? I haven’t used it so I don’t know if it’s automated like that.

Personally I just have yt-dlp in a script run periodically by cron.

Is LinkedIn even useful to you? Every time I’ve gone job searching, I’ve found a bunch on indeed and barely anything on LinkedIn.

8675309 in your local area code almost always works.

You can probably yt-dlp the playlist.