Upgrading Build, Questions on Migration and Best Practices

  • Hi all,


    I built an OMV install out of an old HP 550-150qe. This came with the following items I'm carrying forward:



    • Intel i5 - 4460
    • Hynix 8 GB (2 x 4GB) PC3L-12800
    • Western Digital - Red 2 TB 3.5" 5400RPM Internal Hard Drive
    • Western Digital - Red 2 TB 3.5" 5400RPM Internal Hard Drive


    I run plex and virtualbox headless on there currently (use VM for pihole). I outgrew the motherboard when I purchased 2 X 2 TB WD Red drives a few months ago. I wanted to upgrade further with some 8 TB drives since I'm coming close to capacity (1.5 TB) and obviously ran out of SATA ports on the mobo since it only has 3. I also transplanted the motherboard into a desktop style slim case so I'm actually out of space all together. Fast forward to this week and I've made a new build:


    https://pcpartpicker.com/list/4zKs9J


    So the build above will increase my storage quite a bit. The current setup is that I have software RAID (BTRFS underlying) on the 2 X 2TB drives and all my shares located there. So the question goes to...what should I do to migrate the data or expand my pools? Should I switch up to ZFS? Should I add union and switch to mergerFS? What would you do?


    With the new system, i'll have a M.2 drive to install the OS to and 6 SATA ports. I have a total of 4 WD Reds and I have 2 x 1TB white label NAS drives not currently in use.

  • The current setup is that I have software RAID (BTRFS underlying)

    If you're talking about a btrfs raid1 and are fine 'wasting' 50% disk capacity for redundancy you can simply remain with this setup, add the other disks and rebalance your btrfs raid1 each time. In my opinion the less disks spin the better so I would try to end up with only the two 8 TB disks spinning. With btrfs it's again easy: add one 8TB disk to your pool, rebalance, remove one 2 TB, rebalance, add the other 8TB disk to your pool, rebalance, remove the other 2 TB, rebalance, done.

  • If you're talking about a btrfs raid1 and are fine 'wasting' 50% disk capacity for redundancy you can simply remain with this setup, add the other disks and rebalance your btrfs raid1 each time. In my opinion the less disks spin the better so I would try to end up with only the two 8 TB disks spinning. With btrfs it's again easy: add one 8TB disk to your pool, rebalance, remove one 2 TB, rebalance, add the other 8TB disk to your pool, rebalance, remove the other 2 TB, rebalance, done.


    Well I wanted more of a best practice with multiple drives. I use OMV mostly as a file share for media files and I run Plex from it to serve up the media. I have 2 shares that have a lot of downloads from my torrent machine go to it and I have 4 shares that have very little changes happen in them.


    So what would be the best practice for multiple disks accessed daily with small and large file creation/transfer or is this question to vague?


    I can make the determination for RAID setup at work all day long...but I don't run a media server there :)

  • I've elected to go with mergerFS and SnapRAID for my media server. Just seems to be overkill for me to run ZFS or RAID when my overall space will be under 10TB. I'll eventually post back with results.

    • Offizieller Beitrag

    Keeping in mind that video files are large, what will be the bulk of your media storage? Video?


    *Edit: I saw the thread you looked at RE SNAPRAID & UnionFS. If you set up that way, you should be fine.

Jetzt mitmachen!

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