Boot disk broken, how to proceed?

Hello everyone,

after a couple of years of flawless service, my truenas scale boot disk suddenly broke without apparent recovery in sight (it is not detected on the NAS itself when live-booting debian and I connected it to my laptop using an USB adapter I know is working and it does not enumerate).

As you might imagine I don’t have a backup of its contents (shame on me) but I do have the tar archive of the system upgrade I performed a few weeks ago (from 23.10.02 to 24.04.02, filename truenas-TrueNAS-SCALE-23.10.2-20240802154106.tar).

I’m wondering what is the best way to proceed from here when I’ll get a new device to use as boot disk:

  • is the installer going to prompt me to load the tar file at some point? (In this case I’d probably install version 23.10.02, I guess?)
  • should I complete the installation of 24.04.02, import my storage pool and just ignore the tar file?
  • something in between?

I’d very much appreciate every advice.
Regards,
Daniele

Welcome to the forum :slight_smile:

You can install 23.10.2, import your config after the installation then upgrade to 24.

The installer will not prompt you for the tar file but you can do it from the web ui afterwards.

I can’t add directly to @MSameer 's response other than to say it is a quick and straightforward procedure which causes me unnecessary stress every time I have to do it!

Once you have done it a couple of times, it becomes almost second nature - just make sure you have an up-to-date config at all times. There is a brilliant script called multi-report which, amongst other things, creates automatic configs and sends them to you by email which you can set to run periodically (e.g. once a week).

I spent some time looking into making a second “spare” boot disk which I could keep safe in a drawer somewhere, but in the end I decided that the re-installation process is straightforward enough for me not to bother.

If you don’t have a backup of your system configuration file then it may be a bit more complicated - ask again here and we can help you,

I use a mirrored boot pool in addition to keeping a backup of my config (in the dara pool though).

I just prefer to not have to act immediately if a boot drive fail and redundancy helped me testing core to acale migration.

I must admit that this aspect still attracts me, knowing the feeling of dread which ensures when something goes wrong. Having an immediately available boot drive is still something I am thinking about.

The trouble is, in my particular situation, I’ve run out of SATA connections and additionally I am not sure how my BIOS would react when one boot drive fails and the other is still present in a mirror.

In any case, keeping a spare single drive is still an option for me to think about.

You can keep a spare boot drive and a recent enough backup and be happy.

Just clone your boot to the spare boot drive via a sub enclosure.

I just have spare ports so i use them.

My TN froze suddenly becauae the boot SSD barfed. If a hard reset does not bring the server back up then IPMI can be used if you have it. Of course manual intervention will be needed but I don’t want to go out on a saturday night searching for a shop just because my server is down.

Thanks everybody for the warm welcome and the helpful replies!

@Protopia I’ll first try importing the configuration archive I downloaded during the last system upgrade and as for further guidance if this does not work :slight_smile:

I’m waiting for the replacement and will follow up with the results. Thanks again!

Disk arrived and everything worked as you described. Many thanks again!

2 Likes