1

Today I was trying to find out why my server stalled after certain amount of requests on apache2, but couldn't figure why, so I decided to upgrade it(apt-get update & apt-get upgrade) before I gave my server a restart due to the fact I found it easier than killing the processes

I hadn't booted the server since I got it three months ago(FYI the server is from a German company, Hetzner) and to my surprise it didn't accept my key anymore nor my login credentials, after tearing my hair from my head I somehow ended up trying to login with the original login info I received and to my surprise it all worked too well

But I wasn't greeted with the same environment I had became used to, it was like the initial installement of debian, the server name and all, at this point I almost broke in tears, such frustration

I was googling up and found some promising results regarding faulty hard drive and RAID configuration being messed up and someone telling the host companys support was good, but after checking the raid configuration and health of the drives they all looked good

But the files are at the state they were when I received the server initially

How can this be? What could of happened? Only thing that comes to my head all this 3 months I've been ran from some virtual drive or something? And no changes were committed to the HDDs whatsoever? But I can't see how they're persisted now after numerous of reboots?

The other reasonable answer I can think of is that I had written all the data on the other drive and somehow the RAID failed to write on the other drive and BOOM now after reboot or something.. it decided to rewrite the drive that had no changes persisted on the other drive?

Or what could of caused such behavior after all, I am now hesitated to just start over, but I don't know how to deal with this if it happens again? I didn't have too complicated setup after all, but this kind of behavior just makes me lose my confidence even in managing a single small server

Update:

I decided to see if there's any data left for me to build upon and started opened up in the server providers rescue mode and stopped the array I was running and mounted the both drives separately on different places and checked and they both were as expected on the same state as they were initially, so basically good bye hours spend configuring postfix and so on, at least I am a experience richer, that doesn't really help me back on my foot with the question what really happened

I guess the only solid answer is that I somehow had my data written into one of the disk and after a restart it mirrored it to the other disk and boom, everything gone.

Improve this question

1 Answer 1

Reset to default
1

This link may provide some assistance in establishing where your machine is at after the upgrade https://unix.stackexchange.com/questions/79050/can-i-rollback-an-apt-get-upgrade-if-something-goes-wrong

Apt doesn't make rolling back a simple operation and the first thing would be to identify a list of all the packages that were upgraded.

I think the key takeaway from this is to not just plow into doing an upgrade before you get a solid idea of the issue you're experiencing. In the future definitely consider taking some sort of snapshot before doing anything drastic (and regularly for backups).

Improve this answer
4
  • Thanks for your input, however the problem I am having is not that I would want to go back to the same state before upgrade, but that my data(other than packages) were destroyed after an restart, and I don't think that had anything to do with the upgrade, at least not a single time I have performed an upgrade have my OS silently been reinstalled.
    – Ruuhkis
    Apr 16, 2016 at 23:51
  • Is this a VPS or a physical box?
    – swathe
    Apr 16, 2016 at 23:57
  • nvm I see it's a physical one. What's the RAID configuration on the drives? If it's something like raid 1 after backing up the volum maybe try booting from each drive by itself and see if there are any changes but I'd first install some RAID monitoring software and see if there are any errors. If the errors are SMART related the controller might be disabling the drive, and you may be able to turn off SMART checking on the controller to allow it to boot.
    – swathe
    Apr 17, 2016 at 0:05
  • virtual RAID1 using mdadm with two drives, I am going to do the checking asap, I am not really familiar with raid technologies other than small home backup solution, but this came with the virtual one out of the box
    – Ruuhkis
    Apr 17, 2016 at 0:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .