Toshiba drive dead - ata exception Emask 0x0

1. offizieller Beitrag

    Hello,


    in my NAS with OMV 6 I have a "Toshiba Enterprise Capacity MG08ACA 16TB" for storing my data. This drive isn't usable anymore. I found in syslog:


    The error occured on saturday morning. I register at saturday evening, that the NAS has problems and reboot - without success.


    What can I do, to get access to my data on this drive?


    Thanks in advance


    Matthias

    I can run smartctl:

    • Offizieller Beitrag

    You could try fsck -t EXT4 /dev/sda this will take some time, due to the size of the drive and it will ask if you want to fix each error man page


    fsck is a linux check and repair utility for linux file systems, this should work but it could also result in some data loss


    From your smart output #5 is showing 81 this is one alarm bell, add to that #196 is showing 71 and #197 1

    I cannot execute fsck on the running OMV, because /dev/sda1 is in use. But I start a Manjaro Live System from usb (I use Manjaro on my desktop) and execute:

    Code
    sudo fsck -t EXT4 /dev/sda                                                                     ✔ 
fsck from util-linux 2.38.1

    Then I try:

    So, a simple fsck don't works.

    • Offizieller Beitrag
    Zitat von matkoh

    I cannot execute fsck on the running OMV, because /dev/sda1 is in use

    ?( that would suggest the drive is mounted which makes no sense from your first post


    Zitat von matkoh

    But I start a Manjaro Live System from usb

    Are you certain that /dev/sda is the actual drive given you are now booting from a live distro, blkid or lsblkid would confirm


    Zitat von matkoh

    So, a simple fsck don't works.

    Once you confirm that the drive is /dev/sda try fsck on the partition should be /dev/sda1

    Zitat



    ?( that would suggest the drive is mounted which makes no sense from your first post

    It is strange: In OMV in drives I see the Toshiba drive, but in filesystems not. And I hear, that the drive is accessed the whole time - when start in OMV and when start in live-system. There is a stepping sound....


    Zitat

    Are you certain that /dev/sda is the actual drive given you are now booting from a live distro, blkid or lsblkid would confirm


    Zitat


    Once you confirm that the drive is /dev/sda try fsck on the partition should be /dev/sda1

    That seems to work:

    Code
    sudo fsck -t EXT4 /dev/sda1                                                                  8 ✘ 
fsck from util-linux 2.38.1
e2fsck 1.46.5 (30-Dec-2021)
/dev/sda1: recovering journal

    It is still running - till now without any more messages. I tell, when it is finished.


    Matthias

    In the first hours I get this messages:

    and more of the last messages. Now, on friday morning, it is still running, but in the last days I get no new messages.

    I tell, when it is finished.

    • Offizieller Beitrag
    Zitat von matkoh

    Can you guess, how long it takes to finish the check / repair

    No, but I did look at the man pages and there is an option you can add to display a progress bar, but according to this it can also be done whilst running

    There is also information in the e2fsck man pages under the section -C fd

    • Offizieller Beitrag
    Zitat von matkoh

    I try, but I can only see, how long fsck is running, not how much of the disk is checked.

    TBH I have no idea what it does, I found those links which say how a user can activate a progress bar, none of them state what it actually displays

    I cancelled the fsck-process with CTRL+C while hoping, the drive is ok now and get

    But when I start the NAS normal, there were many errors about the drive. So I start with manjaro-live-usb again and execute

    So, the check is at 23.5% of the drive.


    Meanwhile I have more messages "Inode xxx was part of the orphaned inode list. FIXED" and it shows me 40.1%


    It seems to be a different, if fsck is executed with "-C 0" at start or if later a "kill -USR1" is used.

    • Offizieller Beitrag
    Zitat von matkoh

    It seems to be a different, if fsck is executed with "-C 0" at start or if later a "kill -USR1" is used

    I think we've both learnt something here, but that was seriously ballsy to stop the running fsck, I've read elsewhere that that could corrupt the file system

    I've only used fsck once, but that was on a small drive but I was able to recover nearly 100% of the data, I was lucky. But many years ago I invested in Spinrite and that is my usual go to for recovery and repair

    Yes, it was no good idea...


    I have backups with rsnapshot on a usb-drive, but where to restore? The Toshiba drive is only 3 months old, there is garantee. But I don't want to send a disk with my data...


    Now the fsck is finished with this messages:

    Another attempt to start fsck gives only the version, nothing more.


    I just start the NAS, but still get errors:

    Code
    ata1.00: exception Emask 0x0 SAct 0xffffffff SErr 0x0 action 0x0
ata1.00: irq_stat 0x40000008
ata1.00: failed command: READ FPDMA QUEUED
ata1.00: cmd 60/80:c8:00:08:c0/00:00:15:00:00/40 tag 25 ncq dma 65536 in
         res 43/40:80:20:08:c0/00:00:15:00:00/40 Emask 0x409 (media error) <F>
ata1.00: status: { DRDY SENSE ERR }
ata1.00: error: { UNC }
I/O error, dev sda, sector 364906528 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

    and so on....

    • Offizieller Beitrag
    Zitat von matkoh

    I have backups with rsnapshot on a usb-drive, but where to restore

    Unless you have another drive, you can't use that Toshiba, if that's a complete backup of everything on that Toshiba you could try running dd to zero that Toshiba before returning it.

    If you do a search there is also a ddrescue option (I've not used it) but in effect you create an image of a drive, but to do that you need a drive to write the image to

    My plan is to buy another drive - maybe the same Toshiba drive (or better another model?) - and not create a RAID, but use one drive as main and rsync the directories to the other drive - maybe once a week. So I can fast switch the shares if the main drive fails again.


    But at first I want to repair the Toshiba drive to copy the whole data to the usb-drive. There is enough space to do this.


    After this I try work with OMV as usual and see, if again errors occur. Then I will return the Toshiba drive for repair or replacement.


    For now, I start with Manjaro-Live-USB and start another repair. Here the first messages:

    Yes, but when answer with 'a' instead of 'n' or 'y' all following questions are answered with yes, so no further interaction is needed.


    Do you have a opinion about a new drive? Is it reasonable to buy a drive, which is crashed after short time? When I search for meanings to buy drives, everyone prefers another producer and model.

Jetzt mitmachen!

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

Benutzerkonto erstellen Anmelden