disk fs corruption since core linux 7

    Hi,


    I have 5 disks in fuse.mergersfr

    I work Docker via portainer et images of deluge and plexmediaserver.

    Everything was working until the core of linux went from 6 to 7.

    After that, at each reboot, I lose 2 big disks not mounted at startup and of course mergerfs.

    I run fsck -y /dev... I recover my disks then restart mergerfs, start dockers and everything works until the next reboot.


    I don-t understand what changed between linux core 6 and 7

    Gemini told me that systemd took over and it seems that that docker still runs whereas my disk are unmounted.

    I really don't know what to do. For the moment I run OMV with linux 6 core.

    Any idea? thanks

    Over the weekend when I rebooted my NAS, I had one of the fs on one of the ssd broken (I have 2 ssd and 3 hdd).

    The NAS rebooted normally, no errors observed during reboot, but there were errors during mounting of that fs.

    I also had to do a fscheck to fix it, but I lost approximately 18 gb of data (not important data since the data was replicated from another drive).


    It was the first time a fs was corrupted 3.5 years since I setup the current NAS configuration.

    In my case, I would not blame the kernel, I think that it was a different problem.

    with kernel 6, it doesn't happen. As soon as I reboot on kernel 7 it happens. So I decided to stay on kernel 6 for the moment.

    Gemini says that systemd operates booting or shutdown in parallel operations so Docker may continue to write when disks are unmounted.

    I have no clue why it happens to me. I doesn't seem that it is a problem for others.

    If someone knows how to force the system to wait the proper stop of deluge in docker and docker shutdown before unmounting disks, that would be nice.

    Thanks

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!

Register Yourself Login