After upgrading from 24.04.2.3 to 24.10.0 after the “Welcome to Grub!” prompt but before the actual Grub menu, I started getting these weird “unknown device” and “compression algorithm 34 unsupported” that I don’t see on any other of my servers running Linux, specifically Debian, and didn’t see on the pre 24.10.0 upgrade.
Hardware is a Dell PowerEdge R440 in my home lab. Hardware outside of standard is a Nvidia Quadro T400 and Broadcom BCM57810 Dual 10Gb SFP+
I can’t imagine I did since I have no idea what that even is, I’m an IT pro and don’t generally just click things for fun.
So, if I backup the config and re-install, what has to happen after restoring the config to get apps, VMs, and all the shares back up and going? I know it’s a backup and restore, but I’ve come across times when there’s still a ton of work afterwards to get things going again.
In my experience of doing it like 10 times now; nothing at all needs to be done additionally.
Honestly I should just perform a clean install every upgrade because random crap gets fixed every time I do a clean install & config import vs random issues when I just do a regular update.
That is unless you have some not recommended things like partitioned boot drives - in which case good luck & hope you figure it out.
[EDITED]
After re-reading your post and ignoring the annoying GIF, I get your point that it works without concern.
I agree that sometimes, it is much better to perform a full install rather than an upgrade, but sometimes it’s not practical to do that. If TrueNAS only did ZFS on the boot pool, then I could see them moving to that and doing just an import/restore similar to how things like ESXi does it.
In this case though, having the boot snapshots to revert back to is a bonus and would be lost in the rip and replace upgrade method, so there’s something to weigh in. In this case, that’s been a saving grace because going from 24.10.0 to 24.10.0.2 causes my apps service to not even try to start, so it’s nice to have that “easy button” of booting back into the previous boot environment. The rip and replace would’ve really caused me a late night an headache there.
Anyway, what is the risk to my system if the restore does not work as expected, can I reattach my storage pool and go through the hassle of recreating the VMs and apps?
Lastly, how do you all get the “See My freeNAS system” item after your posts?
I am hoping the following posts good regarding the ‘See My FreeNAS’. I hope this preformed text helps you try it out with the second section what it appears to be on forum posts
we have, it’s called details.
[details="Summary"] This text will be hidden [/details]
Summary
[spoiler]current spoiler is old inline spoiler[/spoiler]
I haven’t done a restore with apps, etc. The separate boot pool and reloading the saved config is supposed to restore everything. On simple setups of a single boot drive (no mirror). I can completely replace my boot drive, Install TrueNAS and reload the configuration and get back my data pools along with my SMB shares that were set up.
If you are worried about doing it, you could use a new device for booting and then do the new install and loading the config. That way if there was a problem, you just power off and put the old boot device back and you would be back to where you are now.
To sum up what they are saying is everything should just work as long as you didn’t deviate from a standard TrueNAS install. Some don’t like using an entire device just for the boot and will partition it to use it as storage too. That is where troubles start.
Defiantly was careful to not allow the general storage pool to exist on the boot pool disks. The System Dataset Pool is on my Pool1, which is not by boot pool.
I can try swapping out the disks, at least pulling both of the current boot disks and just placing a new one in. If it works, then I can expand the boot pool to a new mirror. Probably wouldn’t be bad so that I can go with a smaller disk, because 480Gb is probably way more than I will ever need for a boot pool unless I want to save a snapshot of every upgrade. Maybe replace with a couple of Intel SAS disks with PLP.
The re-install did work, getting rid of the errors with the benefit of switching out the SSD’s with datacenter one’s that have PLP and are a little faster and are SED if I decide I need that.
