Bug ID | 1233698 |
---|---|
Summary | btrfs child eb corrupted |
Classification | openSUSE |
Product | openSUSE Tumbleweed |
Version | Slowroll |
Hardware | x86-64 |
OS | openSUSE Tumbleweed |
Status | NEW |
Severity | Normal |
Priority | P5 - None |
Component | Basesystem |
Assignee | screening-team-bugs@suse.de |
Reporter | dzmitry@kanunnikau.online |
QA Contact | qa-bugs@suse.de |
Target Milestone | --- |
Found By | --- |
Blocker | --- |
I've been having some issues with my filesystem. Noticed some errors, so over the weekend tried to fix them: - used latest OpenSuse Leap live USB to run 'btrfs check', - once check completed in my infinite wisdom decided to run 'brtfs check --repair', ignored the warning about only doing that if someone knowledgeable told me to do so, - rebooted into the installed system, ran 'btrfs scrub'. It ran, but complained about "cache and super generation don't match, space cache will be invalidated". Everything seemed fine then. This morning I try to boot the system, but it would drop into rescue mode with "filesystem readonly" message. I ran the same live USB and started following this guide (https://en.opensuse.org/SDB:BTRFS#How_to_repair_a_broken.2Funmountable_btrfs_filesystem). There I arrived at the notice "the situation is serious enough to justify a bug report, please!", so here we are :) The partition (which lives in a LUKS container, to make things more interesting) refuses to mount, so now I am at the 'btrfs restore /dev/mapper/cr_root /...' stage. Restore seems to be working, it has copied close to 200GB from the unmountable partition. Yet the whole system froze at some point! The live USB system froze completely, didn't respond to mouse or keyboard. The screen was on and the clock was stuck at the time about 40 minutes ago until I hard rebooted it. I have had similar issues with the installed system: - i would try to use pandas to parse a 5GB CSV file, - at some point the system would start stuttering (music playing would stop/resume, mouse would jump instead of moving smoothly), until it would freeze completely, - leaving the system overnight would result in a frozen system in the morning, with clock stuck at somewhere at night, hours ago. Using shell scripts to parse the same 5GB CSV file would work fine, so the file seemed not to be corrupted. I suspected that pandas would result in too much swap use, which somehow messed up the system, but I haven't investigated that yet. Now I saw a similar thing with the live USB system, when it froze completely while doing 'btrfs restore'. I hard rebooted the system, booted into live USB again, tried 'btrfs restore' again. It skipped previously restored files and this time was able to complete the process. Could it be hardware failing? What should I do now? Continue with the repair guide (https://en.opensuse.org/SDB:BTRFS#How_to_repair_a_broken.2Funmountable_btrfs_filesystem)? Try to debug the issue more in depth in case this turns out to be interesting to OpenSuse developers? Thank you