OMV6 /dev/sd*

  • Hi There,

    I manually altered and prepared my HardDisks. I even wrote /dev/sda, /dev/sdb etc on the labeled hard drives.


    Then I put every HD in the machine with a gap of 10 seconds.
    The result is this...
    How can I change this labeling done by OMV6 and let it correspond with my preparations?

    The reason I ask is because I know what is on /dev/sda or /dev/sdg etc.

    Even though sda and sdf where last to inserted in the PC they aren't last in the list.

    Kind regards,
    Guy Forssman


  • When you mount the file systems, they are going to mount by UUID anyway, not by label.. so the drive letters, labels, etc.. are irrelevant. You essentially spent a lot of time doing this for nothing,. I'm a bit of an organization freak myself, and use symlinks to try and make sense of the madness. This really doesn't effect anything other than command line access and and my volume paths for docker-compose and stacks. I very rarely look at the absolute paths of my drives, so it's really not a huge deal irregardless once you get everything setup.

    Air Conditioners are a lot like PC's... They work great until you open Windows.


  • Even though sda and sdf where last to inserted in the PC they aren't last in the list.

    This is how the Linux kernel names devices and that is one reason why OMV does not use that device names internally, because they are not predictable. The Linux kernel may name a device different on every boot.

  • but... you can see labels on webGUI:



    And of course you can label it by shell,



    as said in previous post, do not label sdx, Y use HDx instead, is enought to identify for me

  • Don't change the names of the disk inside OMV, just put new labels to your disks with their ID.

    That is how I do it, because I do not want to pull the wrong disk when the system is running.
    Yes this means I have to put an a new sticker when I change disks, but this happens every six years or on disk failure.

    If you got help in the forum and want to give something back to the project click here and write up your solution for others.

  • Thanks all for the very valuable input.

    Because I'm on OMV6 and experimented a lot I pulled all drives and did a fresh install.

    Now of course OMV6 had no choice than to follow the order I was putting in the drives.
    Not something you want to do everytime indeed.

    It's true these drive names can change on reboot but since OMV6 in the gui itself is still using and that's what showed up by default in filesystems. I really wanted the names correct so I knew what's on it.

    And then I found thanks to raulfg3 a little window on the upper right side and this is the result.
    Thank You all....

  • It still doesn't matter.. they may randomly change at some point. I did a clean install of omv 6 yesterday... When I started, my OS drive was sda... It's sdc now. In fstab, the drives are mounted by UUID, not /dev/sdX. As votdev said above.. it may rename the device every single boot.


    Drive labels would be the easiest way to effect this, as said..

    Air Conditioners are a lot like PC's... They work great until you open Windows.


  • It's true these drive names can change on reboot but since OMV6 in the gui itself is still using and that's what showed up by default in filesystems. I really wanted the names correct so I knew what's on it.

    OMV is using the non-predictable device files in the UI because they are shorter and more meaningful for the normal user. I don't think you'll be happy when there is something like this /dev/disk/by-id/scsi-SATA_ST3200XXXX2AS_5XWXXXR6 or /dev/disk/by-uuid/ad3ee177-777c-4ad3-8353-9562f85c0895. Under the hood ONLY predictable device files are used.

  • OMV is using the non-predictable device files in the UI because they are shorter and more meaningful for the normal user. I don't think you'll be happy when there is something like this /dev/disk/by-id/scsi-SATA_ST3200XXXX2AS_5XWXXXR6 or /dev/disk/by-uuid/. Under the hood ONLY predictable device files are used.

    That is indeed almost unreadable.
    And indeed these device name can change on reboot.
    But OMV6 doesn't need to use the real /dev/sda1 in the file system on the Gui.
    Its feasible that OMV6 gives for understanding a sda name to a ad3ee177-777c-4ad3-8353-9562f85c0895 disk.
    And when this is only a mapping the user can rename the sda to whatever he wants.
    Internally OMV6 is using the UUID so that never changes.

  • I would appreciate if one could give a name to a disk, independent of the sdX name.

    It could be mapped to the UUID of the disk and be manually maintained in disks / edit.


    Even more because if you use disks in a zfs pool they are still considered unused in omv and in the drop down fore wipd.

    If you got help in the forum and want to give something back to the project click here and write up your solution for others.

  • I run OMV5. Every drive that isn’t boot can be wiped. However when trying to create a RAID they are not available. So it marks the drives as used in some instances. However you still need to watch what you are doing.


    OMV is free and designed for functionality. If you want something that doesn’t need some technical knowledge then probably better to buy a Synology or QNAP.

    Former Xpenology user moved to OMV 5.x with ZFS.

    HP Microserver Gen8 - 16GB RAM - 1x 32GB USB - 1x 480 GB SSD - 4x 16TB Exos (Shucked) / ZFS - OMV 5.x bare metal

    HP Microserver Gen7 N54L - 8GB RAM - 1x 32GB USB - 1x 240 GB SSD - 4x 4TB / ZFS - OMV 5.x bare metal

  • This is what i mean


    Not very user friendly, espcially, if you have more drives of the same type.
    And the later two are looking different to the former two, because they are passed through ESXi RDM disks and not connected to the HBA passed through to OMV, otherwise all disks would look very similair.


    But maybe you are right and I should go for something which fits my technical skills :-)

    If you got help in the forum and want to give something back to the project click here and write up your solution for others.



  • I didn't want to make you feel bad..I'm a beginner here and of course my post make more sense to me than anybody else.

    When I read your post indeed this is what I was meaning to say. For clarification we can rename the long UUID to whatever we want.
    some will call it c,d,e etc others will use sda,sdb,sdc or Disk1,Disk2 or 1,2, you get my drift.
    Strange that zfs disks arent recognised has having data and can be easily wiped. I assume you'll have the zfs plugin.

    Kind regards,
    Guy Forssman

Participate now!

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