Missing FileSystem after RAID5 problem

  • Hello,


    I have been browsing and trying several solutions on this forum without success, so I need your help...


    Prerequisites: OMV4, HP Z420 machines with builtin raid, 3HDDs of 4TB. One HDD failing.


    At a certain point the raid fall. In GUI no RAID displayed, missing filessystem and only the 3 HDDs visible in Disks tab.


    2 raid appeared in mdstat in cli, one with 2 HDDs and one with the third one. I stopped them both and create the old with:


    Code
    mdadm --verbose --create /dev/md127 --chunk=512 --level=5 --raid-devices=3 /dev/sda /dev/sdb /dev/sdc

    After this point I had the RAID5 back in GUI, and after about 400 minutes it shows as clean but Missing files system because of a different name.


    OMV has sent the following message: "

    Code
    Status failed Service mountpoint_srv_dev-disk-by-label-storage
    Description: status failed (1) -- /srv/dev-disk-by-label-storage is not a mountpoint

    I noticed that the name is changed and I tried to rename it but instead of having "storage" in OMV GUI is "office:storage", including the hostname of the server.


    I tried to mount it manually but is says that is not ext4 as it should be.


    Bellow are the requested info on this kind of problems:


    Thank you and advance!

  • Diomet

    Added the Label OMV 4.x
  • This not good, recreating an array is an absolute last resort

    I found that from another reply of yours on another topic...

    This should correct itself with mdadm --readwrite /dev/md127

    Yes, solved. Thank you!


    I still have for the automatically mount:

    Code
    status failed (1) -- /srv/dev-disk-by-label-storage is not a mountpoint

    I tried to do it manually and:


    Code
    ~# mount /dev/md127 /mnt
    mount: wrong fs type, bad option, bad superblock on /dev/md127,
    missing codepage or helper program, or other error
    In some cases useful info is found in syslog - try
    dmesg | tail or so.

    Nothing in dmesg though.


    Also, no luck on efsck.


    It's a "bye bye" situation, isn't it?

  • I found that from another reply of yours on another topic.

    I can only recall suggesting that once with the caveat that it's a last resort.

    mount /dev/md127 /mnt

    The mount point for OMV is /srv


    Is the name displayed in mdadm.conf? Also as per ananas what's the output of blkid

    It's a "bye bye" situation, isn't it

    Not yet

    I tried to rename it

    How?

    Raid is not a backup! Would you go skydiving without a parachute?

  • Is the name displayed in mdadm.conf? Also as per ananas what's the output of blkid

    No, it is not.

    The name should be "storage" and mounted "by label"

    Code
    ~# blkid
    /dev/sdc: UUID="8a1e7d73-19e5-3531-f65a-f4dc1d377bdb" UUID_SUB="8b7e01e0-2e86-3632-5f28-a8f9362ae701" LABEL="office:storage" TYPE="linux_raid_member"
    /dev/sdd1: UUID="48A1-D254" TYPE="vfat" PARTUUID="554f2b97-5a95-4484-8f4b-709396ac7666"
    /dev/sdd2: UUID="5c724eba-2d04-4cac-bfd2-638f9f55c0b2" TYPE="ext4" PARTUUID="963556cf-f1de-4833-a661-fb87f9b7df75"
    /dev/sdd3: UUID="0d39422a-174e-45e7-a9eb-f496146e3652" TYPE="swap" PARTUUID="4e3a30fc-ed2b-4508-8b75-b9d2dfb37c96"
    /dev/sdd4: UUID="3247bd9e-62d0-4fe0-a52d-1a71e4a8951d" TYPE="ext4" PARTUUID="397b356a-4c87-423d-8276-a04771664ea5"
    /dev/sdb: UUID="8a1e7d73-19e5-3531-f65a-f4dc1d377bdb" UUID_SUB="4ff6b63d-4587-e513-7f21-f0ea60a0d609" LABEL="office:storage" TYPE="linux_raid_member"
    /dev/sda: UUID="8a1e7d73-19e5-3531-f65a-f4dc1d377bdb" UUID_SUB="125a8656-c4d0-fa91-303a-fdee59e467db" LABEL="office:storage" TYPE="linux_raid_member"


    How?

    Code
    mdadm --assemble /dev/md/md127 --name=storage --update=name /dev/sd[abc]
  • mdadm --assemble /dev/md/md127 --name=storage --update=name /dev/sd[abc]

    That's how I would have done it but without the md so just /dev/md127


    Have you tried ananas suggestion but with /srv so mount -t ext4 /dev/md127 /srv


    The blkid shows no reference to the raid, have you done an --examine on each drive mdadm --examine /dev/sda

    Raid is not a backup! Would you go skydiving without a parachute?

  • I have made the suggested changes and the output is:

    Code
    ~# mdadm --assemble /dev/md127 --name=storage --update=name /dev/sd[abc]
    mdadm: /dev/md127 has been started with 3 drives.
    Code
    ~# mount -t ext4 /dev/md127 /srv
    mount: wrong fs type, bad option, bad superblock on /dev/md127,
    missing codepage or helper program, or other error
    In some cases useful info is found in syslog - try
    dmesg | tail or so.
  • That's odd the name change has appeared to work, but the drives still display office:storage, have you checked mdadm --detail --scan --verbose and mdadm --detail /dev/md127

    Raid is not a backup! Would you go skydiving without a parachute?

  • Now I have in dmesg:

    Code
    [67255.849584] EXT4-fs (md127): VFS: Can't find ext4 filesystem

    And the requested output:

  • EXT4-fs (md127): VFS: Can't find ext4 filesystem

    :( that would suggest you've toasted it by using --create as per your first post. You'll have to rebuild, but I would suggest a clean install, it's quicker and a lot less hassle, but I would secure wipe those drives.

    Raid is not a backup! Would you go skydiving without a parachute?

Participate now!

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