Before panic (I need urgent help !)

  • Hi,
    I have a 8 x TB disk RAID 6 array.
    I have added 2 disk in the case and switched on.
    The I have booted and open the web open media vault GUI.
    I have wiped the 2 new drives and then selected the RAID 6 array and clicked on GROWN. I have the selected the 2 news drives and the GUI became gray. After a couple of minutes I had a "communcation error".
    After several minutes of "communication error" I have been forced to RESET the system.
    Now I have rebooted the NAS and the system stopped at a RESTORING JOURNAL (or something like this) message.
    From time to time I receive a TASK MD127 has stopped for more than 120 sec
    and
    FSCK task has stopped from more than 120 sec.


    No drive leds are moving.
    How do I go on from here ?


    NOTE: I have read on consolle that the array is 10 disks but it never seen the resilvering start. I suppose the configuration has been saved without resilvering being performed.

  • Hi,
    I have started the SystemrescueCD and launched GParted.
    It recognize an array made of 8 disks; even check disk is ok.
    Open Media Vault recognize an array of 10 disks.
    I suppose Open Media Vault stopped after having changed the disk configiration but I don't know if this change have been made only inside OMV config, or it has been made somewhere on the disks to.
    I don't want to remove the 2 disks phisically as I suppose OMV will take wrong countermeasures as it will find 2 disk less than its config.
    At the same time I don't know if the number of this is retured from a command which analizes the RAID setup or it is coded in a config file. In the first scenario we would have SystemrecoveryCD returning a different number than OMV internal commands.
    Also I don't want to reinstall OMW as I don't know what could happen.
    Need some expert advice.


    Regards,
    Giuseppe Chillemi

    • Offizieller Beitrag

    How many drives are actually the array? 8 or 10? If the array is ok in SysRescueCD, it should be ok in OMV now. Have you tried rebooting back into OMV now?

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

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | 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,
    array in System Rescue CD is 8 Disks and it is OK.
    array in OMV is 10 Disks and is reported as Not OK.
    OMV Does not boot. The consolle informs me that task of MD127 is stopped from 120 sec., so it is fsck.


    Regards,
    Giuseppe Chillemi

    • Offizieller Beitrag

    I would boot a Debian wheezy ISO and enter recovery mode.
    cat /proc/mdstat to find the array number. Hopefully, the array is the proper number of drives.
    fsck.ext4 -f /dev/md127 to fix filesystem. Use number from above.
    update-initramfs -u to update initramfs with new array info.

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

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | 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!

    • Offizieller Beitrag

    ISO page. You just need the netinst ISO.

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

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | 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 have downloaded the ISO.
    I have then booted in graphical rescue mode.
    I asked the system to auto detect the array. It created MD127.
    I have then booted on top of the istall shell.
    I have then issued:
    cat /proc/mdstat


    The system recognized 10 drives but as "surprise" I have read RESHAPING 0.3%


    Other Data:
    Level 1.2
    RAID 6
    Chunk 512K
    Algorithm 2
    Speed 11.000K/SEC
    Time left 4300 minutes


    So, speed is about 1/10 of the last reshape
    ... and I have never asked Debian to continue or start this operation.


    What should I do ? Should I let it continue ?



    Also, as far I have undestood until now:


    OMV GUI lost connection after I have started the reshaping from 8 to 10 drive.
    The shell informed me MD127 was not responding in 120 sec for many times.
    I have never been able to take back control of OMV web gui.


    I suppose there was a RESHAPING in progress but without an interface informing me and without consolle reporting it I have "brute force" rebooted. (No hint why web gui was lost communication and why the shell was not reporting the ongoing reshaping).


    System Rescue CD for some reason did not mount a 10 drives array but an 8 drives one. It recognized only 8.
    Debian has recognized 10 and automatically started reshaping.


    Do you read something wrong/dangerous in this situation ?


    Regards,
    Giuseppe Chillemi

    • Offizieller Beitrag

    I would just let it finish and see if it is ok.

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

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | 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!

    • Offizieller Beitrag

    What does that mean? Post 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 7.1.4 | 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!

    • Offizieller Beitrag

    Unforunately, it is reshaping slow and is a very large array. You could try executing:
    echo 50000 > /proc/sys/dev/raid/speed_limit_min
    to see if it speeds up.

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

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | 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,
    I have tried but it has not worked.
    I'll wait the remaining 2 days to let it complete.

    Maybe I have to try to add another SAS cable to my LSI 9211-8i to HP SAS expander to obtain link aggregation (when I will discover how to find HP SAS expander firmware version !)


    Regards,
    Giuseppe Chillemi

  • Final update:
    Reshaping has ended yeasterday night, then started resyncing: another 16 hours.
    Then I have tried you commands:


    "fsck.ext4 -f /dev/md127"


    It reported an error that md127 was in use and mounted.
    I have skipped the following command:
    update-initramfs -u


    And started SystemRescueCD and GParted.


    It checked the EXT4 FS and started resizing and rechecking te FS.


    I have then booted OMV.
    The last 10 minutes where occupied from a unexpected "Cheking Quotas" and finally the OMV PROMPT !


    Everything is there. Nothings seems have been lost


    Thanks for your help.


    A last note about Debian netinst iso. Its "RESCUE MODE" is not so safe. I didn't expect reshaping auto starting after array recognition. A rescue mode should maintain everything stopped until you ask to do something.


    Regards,
    Giuseppe Chillemi

Jetzt mitmachen!

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