RAID5 and EXT4 filesysten status "missing"

1. offizieller Beitrag

    Hi at all,
    I have a problem with my RAID 5 system, I receive EXT4 filesysten status "missing". I try to explain what happened.
    I started my NAS to dowload some pictures and I seen many files are damaged. Immediatly I opened the NAS consolle and I saw one HD was not present. I touched the power cables and I heard one HD was started. I think when the system was off, someone touched one power cable and is go out a little. I restart the system and I read some error in the process starting and when I done a check in the console, under file system I found "EXT4 filesystem status "missing"" and all my data have disappeared. Before posting I read all following links:

    Code
    http://forums.openmediavault.org/index.php/Thread/5543-Filesystem-disappeared-after-RAID-recovery/?postID=50308#post50308
http://forums.openmediavault.org/index.php/Thread/5427-RAID-5-disappeared/
http://forums.openmediavault.org/index.php/Thread/5415-problem-raid-5/
http://forums.openmediavault.org/index.php/Thread/5348-Failed-to-mount-error-on-Raid-5/
http://forums.openmediavault.org/index.php/Thread/4341-Raid5-dissapeared/?postID=37248&highlight=EXT4%2Bfilesystem%2Bstatus%2Bmissing#post37248
https://wiki.sabayon.org/index.php?title=HOWTO:_Repair_filesystem_using_fsck_on_a_raid_setup
http://forums.openmediavault.org/index.php/Thread/4332-fsck-error-8/?postID=37145&highlight=EXT4%2Bfilesystem%2Bstatus%2Bmissing#post37145
http://forums.openmediavault.org/index.php/Thread/3139-Filesystem-Missing/?postID=23446&highlight=EXT4%2Bfilesystem%2Bstatus%2Bmissing#post23446
http://superuser.com/questions/708565/fsck-ext4-unable-to-resolve-uuid
http://howmenet.wordpress.com/2012/07/11/howto-recover-the-problem-of-fschk-exe-unable-to-resolve-uuid/


    My problem is I'm not a expert about RAID technology and I hope you can help me before I doing something of wrong.
    I attached a picture who I taken during the start process with the errors, I hope can help to found the solution.


    Thanks in advanced for everyone can help me.

    Hi guys,
    this morning I'm continue my study to found a solution at my problem and I read others four conversations:

    Code
    http://ubuntuforums.org/showthread.php?t=1854188
http://www.linuxquestions.org/questions/linux-general-1/raid5-with-mdadm-does-not-ron-or-rebuild-505361/
http://www.linuxquestions.org/questions/linux-general-1/failed-raid5-disk-array-questions-about-mdadm-and-recovery-761209/#post3715706
http://ubuntuforums.org/showthread.php?t=1101748&page=2


    I have always more confusion in my mind but I think to search a solution, we need to have more information about the status of the system. I report all I done.



    I checked all drivers and appears to be all ok:


    and the last one:


    I don't know why the command spoken of ext2, my partition was ext4.


    I am hoping someone with more knowledge than myself can help me because I am sure the data is still there, I need only to rebuild the array.


    Have a nice day at all.

    • Offizieller Beitrag

    When an ext4 filesystem loses it journal, it becomes an ext2 filesystem.


    You are missing two drives in your array. That is why it won't mount. What is the output of:


    cat /proc/mdstat

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    Hi ryecoaaron,
    thanks for your reply. I think the response at your question is in the third block of code but I rewrite here to be more comprensible:


    Code
    root@openmediavaultNAS:~# cat /proc/mdstat
Personalities :
md0 : inactive sdc[1](S) sdd[3](S) sdb[4](S) sda[0](S)
      636544279 blocks super 1.2


unused devices: <none>


    Thanks for your support.

    • Offizieller Beitrag

    mdadm --stop /dev/md0
    mdadm --assemble /dev/md0 /dev/sd[abcd] --verbose --force


    It may help to boot into systemrescuecd to execute these commands.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    Hi ryecoaaron,
    at the moment the system is normally started and I can start in recovery mode also. Do you think it's better execute the command booting the system from the CD? Because at the moment my system don't have a CD reader but if need, I search one.


    Thanks a lot for your rapid support.

    • Offizieller Beitrag

    You can try with the system running normally but if it gives you problems, use systemrescuecd.


    You can write the systemrescuecd image to a usb stick and boot from that. I recommend LinuxLive USB creator for writing the systemrescuecd image to usb stick.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    I'll try to do the steps starting the system in recovery mode. I'll know the good tool you reported, but the problem with my system is the MB is too old and doesn't support the boot from USB very well. I execute the commands and I give at you the feedback as soon as possible. Thanks.

    Thanks, thanks, thanks, you saved my data.
    There is only a small problem to solve. The RAID is working only with three driver (sdc,sdb, sdd) and for this RAID is degraded. I tried to add the last driver (sda) but is not displayed in the list when in "RAID Management" I have selected "Recover" option.
    In addition to, before the driver with the OS inside was "sda", now this is signed with "sde", can give problem? I ask because in your command I used "sd[abcd]" and I want to be sure the the fourth driver isn't use because I have written something of wrong.


    Thank you again, I hope to help the community as you.

    • Offizieller Beitrag

    The letters can change. Debian/omv uses the uuid not the drive letter. One of your drives may have failed.


    What is the current output of:
    cat /proc/mdstat
    fdisk -l | grep Disk

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    Ok clear. The output of the command is:



    What means?

    • Offizieller Beitrag

    Just checking to see if the letters were still right. They were. sda is having issues. I would do the following:


    Code
    mdadm --stop /dev/md0
mdadm --zero-superblock /dev/sda
dd if=/dev/zero of=/dev/sda bs=512 count=10000
mdadm --assemble /dev/md0 /dev/sd[abcd] --verbose --force


    If sda still doesn't join, I wouldn't trust it anymore.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    I restarted the system in recovery mode but when I executed the first command I received this error:

    Code
    md: md0 still in use.
mdadm: failed to stop array /dev/md0: Device or resource busy
Parhaps a running process, mounted filesystem or active volume groups?


    I'm searching a way to solve.

    • Offizieller Beitrag

    Probably because it is mounted. What is the output of: mount

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    I think you are right, is mounted, I have this output:

    Code
    ...
/dev/md0 on /media/e8cc0bab-................. type ext4
...


    Do you think I can use

    Code
    umount /dev/md0

    or with the RAID we must to be more carefully?

    • Offizieller Beitrag

    You can unmount it. It may be in use though and not allow you to unmount it.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    Recovering in progress...many many thanks for all again.
    But very early I want to build a new system with most recent components. If you have some good advice, is something pleasing.
    See you around the blog :) .

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!

Benutzerkonto erstellen Anmelden