Pool solution choice?

  • Hi,


    I want to migrate my whs2011 sevrer to a OMV nas.
    I'll use Snapraid plugin, as I'm fully satisfied with its features.


    aufs and mhddfs, seems to be good candidates for pooling.
    I need:

    • No single-file spreading (each disk must be fully readable if the pool fail)
    • Usable (read/write) from NFS AND Samba shares.
    • split-level, as unraid feature (link). The idea is to avoid multi disk spinning when playing a DVD or Bluray Rip.

    Thanks

    • Offizieller Beitrag

    aufs and mhddfs work on top of the filesystem. So, you can definitely still use each disk if the pool fails.


    mhddfs is a fuse filesystem so it can use whatever the underlying filesystem uses. So, no problem with nfs or samba. The only issue with aufs is the lack of support for ACLs but I am against using ACLs anyway.


    aufs and mhddfs are not raid and do not split files. So, if you are watching a single video file, only that drive would need to spin.


    I recommend trying both of them to see which works best for you. You can add and delete pools without worry.

    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!

  • Hi,


    Thanks for the answers.
    In my movies collections, I have a lot of full DVD rips, with their original VIDEO_TS folder and multiple vob files inside.
    I need to make sure that all vob of the same movies are on the same physical disk.
    So I need a sort of split-level as unraid.
    Didn't see anything on aufs or mhddfs

    • Offizieller Beitrag

    If you have all services reading from the pool, the files will appear to be on the same physical disk. I don't recommend using the individual disks for any service.


    I recommend using mhddfs then. The only time files could end up on different disks is if the drive currently being written to is too full.

    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!

  • The only time files could end up on different disks is if the drive currently being written to is too full.


    This could mean that over time movie folders will be non-contiguous and spanned onto multiple drives.
    With other solutions you might be left with storage gaps and wasted space.

    • Offizieller Beitrag

    Yes, movie folders may be non-contiguous but if you are using the pool for everything, it won't matter. You will never see it.


    With mhddfs, you set how much "wasted" space you want. It defaults to 4 gb. Once all drives have less than this setting, it starts writing to the drive with the most free space.

    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!