My SSD suddenly broke - I noticed that Nextcloud gave errors, then saw the PostgreSQL log which said the database had corrupted, and then the SMART status of the SSD warned, to which the SSD is now unaccessible.

My appdata-backup plugin failed to create backups since March somehow, but then I was not even at 10% of having stuff setup, so that is not very usefull.

I had the following containers setup and configured to (finally) work after long time struggles:

bmmbmm01/unraid-smb-password-portal:latest cloudflare/cloudflared:latest

collabora/code:latest

deluan/navidrome:latest

driveone/onedrive:latest

ghcr.io/advplyr/audiobookshelf:latest

ghcr.io/goauthentik/server:2026.2.1

ghcr.io/imagegenius/immich:latest

ghcr.io/immich-app/postgres:16-vectorchord0.3.0-pgvectors0.3.0 ghcr.io/paperless-ngx/paperless-ngx:latest

jellyfin/jellyfin:latest

jlesage/krokiet:latest

jlesage/krusader:latest

lscr.io/linuxserver/duplicati:latest

lscr.io/linuxserver/nextcloud:latest

postgis/postgis:17-3.5

postgres:16

redis:latest

traefik:latest

For me, to setup everything again and get it up and working properly, it'd take weeks again, maybe longer, so I'm desperately looking for help on how to restore the data from my old cache (I bought a new SSD which I already put in the Unraid server) and use it to restore my old configuration.

Also, I would like to know how to prevent catastrophe like this in the future.

Can anyone help me?

Post diagnostics, maybe it's possible to mount the drive read-only.

23 minutes ago, hermanjb said:

Also, I would like to know how to prevent catastrophe like this in the future.

Set up proper backups and regularly check they don't break.

It might not be completely bricked. Try this.

Even if the SSD is completely inaccessible now, it may have overheated or hit a firmware controller crash.

Power down your server completely, unplug the power cable, and let the SSD sit unpowered and cool down for at least half an hour.

Plug the SSD into a Linux machine (or boot into a Linux Live USB) and check if the OS can actually see the drive by running lsblk from terminal.

If the drive shows up as 0GB, uninitialized, or with a generic Phison/Silicon Motion fallback name, the SSD controller has suffered a firmware panic/brick. Software tools won't help in this case.

If it wakes up even momentarily, do not try to boot Unraid normally.

Immediately plug in an external drive and use ddrescue to dump a raw image of the dying SSD before it locks up again.

4 hours ago, hermanjb said:

setup everything again

You can reinstall your containers from Previous Apps on the Apps page, and it will use the same mappings, etc you had before. But without appdata, each application will be starting over.