Can't login in WebGUI, SSH working, root is full

    This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

    • Can't login in WebGUI, SSH working, root is full

      Dear community,

      I am using OMV 3 with SnapRAID and Union filesystem (mergerfs). Everything was working fine.

      I used the command

      mv folder 1 folder 2 in the folder /srv/0b5379ae-d1cb-46a6-abf1-f79b680e4d2d


      After a couple of minutes I got the error that the destination is full. I liked to check what's going on in the WebGUI but I couldn't login anymore.
      In an other post I read that i have to check the root disk with df -h and yes my root disk is full:

      Source Code

      1. Filesystem Size Used Avail Use% Mounted on
      2. udev 10M 0 10M 0% /dev
      3. tmpfs 3.2G 8.8M 3.2G 1% /run
      4. /dev/sda1 106G 106G 0 100% /
      5. tmpfs 7.9G 0 7.9G 0% /dev/shm
      6. tmpfs 5.0M 0 5.0M 0% /run/lock
      7. tmpfs 7.9G 0 7.9G 0% /sys/fs/cgroup
      8. tmpfs 7.9G 0 7.9G 0% /tmp
      9. /dev/sdd1 3.7T 794G 2.8T 22% /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68N32N0_WD-WCC7K4DV8YTA-part1
      10. /dev/sde1 3.7T 3.6T 35G 100% /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68WT0N0_WD-WCC4E5ESCZUN-part1
      11. /dev/sdb1 3.7T 3.6T 86G 98% /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68WT0N0_WD-WCC4E5XX5S5J-part1
      12. /dev/sdc1 3.7T 364M 3.6T 1% /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68N32N0_WD-WCC7K0HRSCKE-part1
      Display All

      With google I can help myself with Linux but I am far from a professional linux user, therefore I don't know how I have to proceed.
      How can I find the data I have to delete on sda1?
      And for the future: How shall I move data on the merferfs "volume" without getting into the same problem?

      Thanks for your help.

      Regards,
    • Morrtin wrote:

      mv folder 1 folder 2 in the folder /srv/0b5379ae-d1cb-46a6-abf1-f79b680e4d2d
      Were you in the /srv/ob53... folder? What is the exact command you used? Moving data on the mergerfs pool wouldn't fill up root unless you were in the wrong directory and moved it to the os drive. What is the output of: du -d1 -BM | sort -n
      omv 4.0.14 arrakis | 64 bit | 4.13 backports kernel | omvextrasorg 4.1.1
      omv-extras.org plugins source code and issue tracker - github.com/OpenMediaVault-Plugin-Developers

      Please don't PM for support... Too many PMs!
    • The Videos folder is definitely on the os drive. The os drive filled so your move command didn't complete. That is why there is still data in that folder.
      omv 4.0.14 arrakis | 64 bit | 4.13 backports kernel | omvextrasorg 4.1.1
      omv-extras.org plugins source code and issue tracker - github.com/OpenMediaVault-Plugin-Developers

      Please don't PM for support... Too many PMs!
    • I started now this command:

      mv /srv/0b5379ae-d1cb-46a6-abf1-f79b680e4d2d/Videos/* /srv/0b5379ae-d1cb-46a6-abf1-f79b680e4d2d/Media/Videos/

      What I don't understand is that the command is still running. It seems the system is copying the data and then deleting it. In windows I am used to it that as long as the data remains on the
      same disk, only the pointer is changed. Does that mean that mergerfs can't handle "move" commands the same way and it is actually copying/deleting the data between disks?
    • Morrtin wrote:

      In windows I am used to it that as long as the data remains on the
      same disk, only the pointer is changed.
      A mergerfs pool isn't a disk.

      Morrtin wrote:

      Does that mean that mergerfs can't handle "move" commands the same way and it is actually copying/deleting the data between disks?
      No because you are copying between filesystems even if mergerfs is working on top of other filesystems. I would've executed this from the drive mount points instead of the mergerfs mount point.
      omv 4.0.14 arrakis | 64 bit | 4.13 backports kernel | omvextrasorg 4.1.1
      omv-extras.org plugins source code and issue tracker - github.com/OpenMediaVault-Plugin-Developers

      Please don't PM for support... Too many PMs!
    • you mean these, or?

      Source Code

      1. /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68N32N0_WD-WCC7K4DV8YTA-part1
      2. /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68WT0N0_WD-WCC4E5ESCZUN-part1
      3. /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68WT0N0_WD-WCC4E5XX5S5J-part1
      4. /srv/dev-disk-by-id-ata-WDC_WD40EFRX-68N32N0_WD-WCC7K0HRSCKE-part1
    • Fun information:

      Move now completed and snapraid diff results looks like this:

      625 equal
      0 added
      197 removed
      0 updated
      74 moved
      197 copied
      0 restored

      The "74 moved" were from the first command which "broke" the system (=> full OS disk). Second move command is recognized as copied and removed. ;)