It’s not the postgres db, it’s the internal pictrs db.

Fix your docker volume permissions so pictrs can write to the folder

Sar?

Nice job!

Just an idea: I bet you could probably squeeze in a disassembled usb hub to make it easier to get those front ports.

Something like https://a.aliexpress.com/_mNBftvQ

You could just swap the two disks and see if it follows the drive or the link.

If the drive, rma it. I don’t put a lot of faith in smart data.

Usually means a failing drive in my experience.

Would be better if you made it “a joke about maize”

No grace period I’m aware of, just removal vs purging. If it’s been removed it can be edited in the db to restore it. Once it’s been purged, it’s gone for good.

Ntfs isn’t going to care or even be aware of the hypervisor FS, zfs or btrfs would both work fine.

Making sure you don’t have misaligned sectors, is pretty much the only major pitfall. Make sure you use paravirt storage and network drivers.

Edit: I just realized you’re asking for the opposite direction, but ultimately the same guidelines apply. It doesn’t matter what filesystems are on what, with the above caveats.

Never attribute to malice that which is adequately explained by stupidity

Had a zfs array on an adaptec raid card. On reboot the partition table would get trashed and block the zfs pool from coming up, but running fdisk against the disk would recover it from the backup.

Had a script to run on reboot that just ran “fdisk -l” on every disk, then brought up the zfs pool. Worked great for years until I finally did a kernel upgrade that resolved it.

I’d believe it. I’ve had hundreds of Linux servers that don’t have any desktop Gui at all deployed on them.

Linux desktop users make up an absolutely tiny fraction of Linux installs.

We’ve definitely noticed an increase in signups at lemmy.ca when that news was announced. Not all active users yet, but a lot of signups

Yes. I’ve always splurged on nice cards for my personal stuff. I think it’s more about the write behavior of Linux than anything else, since I’ve never had a card die in my camera.

I refuse to use a pi with SD at this point. Saving $50 isn’t worth my time to reinstall things.

I couldn’t count the number of failed sd cards I’ve seen across all my fingers and toes.

I’ve seen like 4 ssds in my entire life fail. Plus you could just do mdraid 1 / btrfs across 2 of them if you want

Why not just connect an ssd via USB and save yourself the hassle and torment?

I don’t think there is anything else free. Best you can do is host with someone like ovh that has enough resources to provide basic protection.

What’s your budget?

The internet was designed to route around failure. Taking down an isp upstream wouldn’t generally impact internal routing, or routing between them if they’re peering.

Would be nice if it explains exactly what it does. Right now it’s just a random web app asking for creds…

With the hw MCE errors, it’s probably toast.

You could try reseating or swapping the ram around, if it’s socketed