mergerfs writes only to one disk in the pool

  • Hi guys,


    I have tried on my home boxes omv5 and 6 but I am not able to get working mergerfs on any of them.

    It only uses one of the disks in the pool for storage.

    e.g.:

    /srv/dev-disk-by-uuid-93f25f1c-0ddf-4367-83e6-21010c12623d

    /srv/dev-disk-by-uuid-9f04bc7d-0b92-4f1b-b40e-b4c01715eb74


    Shared folder is set up, writable (all devices are writable).


    When using ftp client to upload data then first disk is filled based on default settings. Uploading stops when 4 GB left on the first disk.

    Then it stops with: error 550 photo_2010-2014.7z: No space left on device


    Other disks are not even used they are completely free and I can write on them via SSH.


    Any hint?


    I have followed several videos on youtube how to set up share, ftp, mergerfs.

    But only Mergerfs does not work for me, other things are working like a charm.


    Br,

    RX

  • You will have to provide us with exactly how you configured the pool.


    Having said that, and knowing nothing more at this point, the best guess is that you chose a creation policy for the pool that includes existing path but the path does not exist on the drive that isn't being written to.

    --
    Google is your friend and Bob's your uncle!


    OMV AMD64 7.x on headless Chenbro NR12000 1U 1x 8m Quad Core E3-1220 3.1GHz 32GB ECC RAM.

  • Hello,

    my set up is omw 5 now with sda for system and sdb and sdc for storage.


    Both disks have EXT4 fs.



    Shared folders set up



    Mergerfs set up.



    As you can see, only one disk is filled with data.


    $ df -h /srv/*

    Filesystem Size Used Avail Use% Mounted on

    mergerdata:062ae1a7-a638-4480-932e-2ca785856d94 29G 11G 19G 36% /srv/062ae1a7-a638-4480-932e-2ca785856d94

    /dev/sdc1 15G 44K 15G 1% /srv/dev-disk-by-uuid-4c5ec9b8-5b67-48f7-af26-f7e3afd08fb0

    /dev/sdb1 15G 11G 4.0G 72% /srv/dev-disk-by-uuid-fd46866d-d40b-41aa-8ce0-474317332a39

    /dev/sda2 456G 2.5G 431G 1% /

    /dev/sda2 456G 2.5G 431G 1% /

    /dev/sda2 456G 2.5G 431G 1% /


    I am wondering what I am doing bad the second disk is not used.




    From FTP client


    Response: 227 Entering Passive Mode (192,168,0,162,141,25).

    Command: STOR debian-testing-amd64-netinst.iso

    Response: 550 debian-testing-amd64-netinst.iso: No space left on device

    Error: Critical file transfer error

  • Good Morning. I have the same problem but changing the method didn't work. I installed OMV on a virtual machine to learn how to configure everything before using it on a real system. I created a pool with two disks (sdb and sdc), called "volume1"; after that I created the shared folder "volume1" and activated SAMBA; inside the folder "volume1" I created a folder called "Downloads". When I copy files to this folder, only disk 2 is used. What is the problem?


    ps: my idea, in the real system, would be to create a pool with 6 disks, with the name "volume1", create a shared folder with the same name and then create the folders that I will use (Downloads, Movies, etc) inside that folder . What would be the best way to do this so that the files I copy to the folders inside the shared folder "volume1" are spread across the 6 disks?

    • Offizieller Beitrag

    What would be the best way to do this so that the files I copy to the folders inside the shared folder "volume1" are spread across the 6 disks?

    This is a policy problem. What policy are you trying to use?

    omv 7.0.4-2 sandworm | 64 bit | 6.5 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.10 | compose 7.1.2 | k8s 7.0-6 | cputemp 7.0 | mergerfs 7.0.3


    omv-extras.org plugins source code and issue tracker - github


    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!

  • This is a policy problem. What policy are you trying to use?

    Well, I tried using both "existing path, most free space" and "most free space". In both cases only one of the disks is being used. I think it's something related to the initial configuration because just changing the method had no effect.

    • Offizieller Beitrag

    I think it's something related to the initial configuration because just changing the method had no effect.

    mergerfs doesn't care about initial config. It doesn't store anything on the disks. If you change policy and don't remount the pool (or reboot), the pool won't be using that policy though.

    omv 7.0.4-2 sandworm | 64 bit | 6.5 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.10 | compose 7.1.2 | k8s 7.0-6 | cputemp 7.0 | mergerfs 7.0.3


    omv-extras.org plugins source code and issue tracker - github


    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!

Jetzt mitmachen!

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