Running Leap 15.0 with KDE desktop. A couple of days ago I experienced strange behaviour with Thunderbird not responding normally to mouse clicks and minimizing. Eventually lost all control and had to recourse to keystroke REISUB to stop system. After re-booting I was presented with emergency texts which I didn’t understand, could not follow and lost as soon as I tried to make progress but the gist of running systemctl gave me some red lines indicating a problem in sda3. This is my btrfs root partition.
I have posted here because the symptoms of the crash which forced my to recourse to keystroke REISUB were then repeated but less seriously on an identical spare machine which was running Debian 9 but with the KDE desktop and this exhibited similar problem which I sorted out by changing to XFCE desktop. Whatever the cause here is my problem:-
I cannot re-boot but get warning messages. All the red lines of text in the systemctl suggest there is a problem in sda3 and the error messages suggest that I might be able to repair the btrfs system.
I booted a live Ubuntu 19.04 and ran Gparted and all looks as it should be. My system has a bunch of drives in a hardware RAID configuration and nothing suggests the fault is with the underlying drives but in my 40GB btrfs root system. The remaining partitions are xfs or swap.
Before I do anything else, how may I mount and read the two data partitions (/home and /multimedia) I believe sda4 and sda5?
PS. Found that only my operating system has been corrupted and high on my list of suspects are the frequent “updates” from KDE. All other partitions are OK. My next course of action awaits advice from those who know more than me but I expect the best course is to reinstall my operating system only this time I shall avoid KDE as I am having too many problems with this desktop.
Time for bed.
Precisely what happens when you try, in detail?!?!
All the red lines of text in the systemctl
What precisely does “in the systemctl” mean? Can you copy some messages here, or take a photo and upload it to susepaste.org? Do you mean journalctl?
suggest there is a problem in sda3 and the error messages suggest that I might be able to repair the btrfs system.
I booted a live Ubuntu 19.04 and ran Gparted and all looks as it should be.
Trouble with booting rarely has anything to do with anything Gparted could help you with. Once installation is complete, partitioning is rarely touched except in readonly mode at booting and mounting. OTOH, trouble with filesystems, which live on partitions (unless LVM is employed), not unusually do suffer impediments to booting, which Gparted can do nothing about.
…nothing suggests the fault is with the underlying drives but in my 40GB btrfs root system.
I can’t provide details, as I never use BTRFS, but the basic procedure can go one of two ways:
- general mode: roll back to an operable state with snapper
- experts-only mode: boot live media that has BTRFS tools and use them to diagnose and repair as necessary
Hi mrmazda, many thanks for the reply. Spent my available time yesterday dusting off my spare machine and bringing it p to date so that I can continue working and then checking that my data had not been fried.
Thanks to live ubuntu dvd I was able to establish that there are no apparent problems with partitions and underlying drive array and my data is sitting safely on the separate partitions.
Sorry I confused terminology. I was working from memory as screen with the detailed error messages which resulted from my booting has long since gone. What I will do now is boot again and take a photo of the screen so it can be seen and I can seek your advice.
It will then take me a while working out how to get this into these pages so if you have any suggestions please let me know, otherwise I could email using my phone which I can do easily!
Hi I have been able to take a picture of the initial emergency mode screen after booting. I have sent the picture to myself by email and it looks fine on my spare machine email. The image is 1.9MB so my next question is how to get it to you. I think it should be available with the link here:-
Please confirm you can read it.
I can now do one of two things, run journalctl or run btrfs check --readonly. The journalctl gave me streams of infor but only two or three red lines and these made it clear to me that all the problems are on my root partition so I am inclined to run the btrfs check and then seek further advice.
I am not having much luck with checking, could it be the partition is still mounted? If so I cannot run from here which leaves me with I dilemma. The only way I can use btrfs commands must be through a live system. If /dev/sda3 is still mounted I a, not sure I can follow the emergency mode instructions unless I override using --force.
Grateful for some btrfs help if anybody is watching!
In spite of all my efforts including using both openSUSE and Ubuntu live distros as well as the emergency system I was not able to sort this out because although the partition was NOT mounted, all the resulting commands I used came back telling me it was mounted.
I have now deleted partition and done a new installation.
The only moral from this is that I must check my root partition regularly to avoid filling it up.
I do believe the system should prevent this with warnings and other mechanisms to prevent this crash.
If I could I would log a bug but I do not have the know-how or time.
The new system works well and I have all my system working because of course my /home was on a different partition. Consider this essential if using btrfs.