I have a problem, I have installed NC on a Ubuntu server 2 years ago (Ubuntu 18.04). So far so good. Now the system crashed, the hard drive (SSD) broke and attempts to be unrepairable. So I have two raids with all the nextcloud files, a system SSD heavily corrupted (actually I have no clue how to access the drive from another (windows) PC to check if at least anything is still readable, in hopes to get config files etc…) and, theoretically, a duplicati backup of the whole system.
But, I am far from being experienced enough to be able to setup a new system on a new SSD and to reinstall the old NC on that or to bring back the whole old thing from the backup drive. Since 18,04 is long overdue I would prefer to use the situation to make it new, and to set up a complete new installation will not be the problem, but to bring back the old NC-system on that must certainly is.
Is anybody here able to tell me what I need to do and how?
Sorry for being that unspecific, but as you might guess, I have no real idea what kind of additional information might be useful (or even obtainable, since I have no access to any old system files)
I tried searching for duplicati procedures but couldn’t find something that would tell me and hope somebody here can and will help me.
Thx for the reply. I assume it will be hard, but that was the first setup as well for me, so I figured if someone can point me in the right direction, I will get it eventually…
And your reply suggests that it will be possible, so that’s light at the end if the tunnel.
I know there are several different raid systems, and I can’t remember what number mine had, but it’s the kind where everything is mirrored to both disks, so if one fails the other one has all the data still. The term mdadm doesn’t really ring a bell, but that doesn’t mean all too much…
And if you say this has only little to do with nextcloud itself, then let me say sorry, I really thought it is a nextcloud problem exclusively (as in how to tell a new NC installation how to gain access to the existing raid system and it’s data…), but if it’s more of a strict Ubuntu problem, maybe I should ask their community…
Do you think it’s possible, because you mentioned doing it manually, to setup nextcloud and then create the existing old users in the database, with password and all? Could that work? Because that sounds fairly simple to me (I know, not THAT simple, but something I can wrap my head around…). To be honest, that was something I thought about a lot already, I just was very afraid that a new nextcloud system would simply erase all the data on the raid disks as soon as you set it up and make it all new. That’s the only reason why I never tried to do it that way.
If you say that can work (and maybe tell me how), I would just begin to setup the whole system to the latest LTS and install Nextcloud and it’s needed environment, after that I just need to know how to tell Nextcloud what to do with the raid disks.
Btw, the raid disks only contain all the nextcloud user data, so their filesystems and user specific Installations like calendars and stuff, nothing else. The Nextcloud software, the configs and the MariaDB as well as php 7.4 were all installed on the Linux system partition on the SSD.
When I will have done the new Ubuntu setup, do you think it would be possible to see if that system can take a look at the old SSD and see if there is still anything readable? Or would that make no sense at all?
Sorry, so many questions. I am no pro here, but I’m a self-tought-person in many things, so I will figure it out if I can get some help from someone who knows the deal. After all I did set up and maintained the original system all on my own. Just that exotic breakdown is more then I can chew. I prepared myself for so many other things that could go wrong, but that’s way out of line for me (ssd’s are thought to be the last thing breaking up, aren’t they?)…
Cheers!
Oke I went ahead and tried a VM installation of Debian 11 with /home setup on raid1. I then replaced / disk and reinstalled. Upon setting up the Partitions RAID was detected. I simply pointed it to /home ( Keep existing data ) and that was that.
You were using RAID 1 as well, this mirror’s disk x and disk y to be identical.
RAID is not specific to Operating System so asking on the Ubuntu forum might get you a bit more help but you can just as well get information about how to setup raid on Debian as Ubuntu is build upon that.
What I’m saying is Your issue is not related to Ubuntu nor NC because you have a hardware failure.
I could help you with creating a video of a ubuntu installation that will recover your raid during installation if that would help. ( It’s gonne be long and boring )
Then after you have a new running system we can look into setting up NC.
Im no expert in Hardware and SSD has been here some time but I’m still old school and have little trust in SSD or NVME. for me HDD all the way. I do have my system run on NVME but I have my system in raid incase 1 fails and got my NC data spread over LVM with weekly backup and daily incremental.
If SSD is the last to fail well apparently not. but such thing are unpredictable.
You could try mounting your broken SSD in your new system afterwards yes.
NC will never delete any of your DATA without exclusive being said so.
And yes this is the way I would go but I think you will lose contacts, calendar, Profile and Shares because there stored in the Database. So uses will have to setup everything from new but will have all there Files.