My 64TB build

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

    • My 64TB build

      Hello all. I've been kinda carefully planning my 64TB build and it is as follows :)

      pcpartpicker.com/list/WCydf8

      I require it to be able to transcode 2 streams comfortably, which is why I chose this processor. Bearing this in mind, is there anything about this build that anyone would change? If so, why?

      Also, I plan on installing OMV onto the 64GB SSD drive. The motherboard has a USB 3.0 port onto it, which I plan on using for the OS drive via a USB 3.0 -> Sata 3 cable. This will leave the 8 sata slots available for the storage drives. Will this work? I've read (albeit the posts were old) that people had problems booting off of USB 3.0 devices and had to use USB 2.0 ports.

      Thanks in advance!
      Case: U-NAS NSC-810
      Motherboard: ASRock - C236 WSI Mini ITX
      CPU: Core i7-6700
      Memory: 32GB Crucial DDR4-2133
    • Dropkick Murphy wrote:

      64TB? 8 Slots...This is a Raid0 with 8TB Drives...
      Is is this, what you really want
      Unless they are pooled with the unionfilesystems plugin. Hope there is a backup planned for this...
      omv 4.1.11 arrakis | 64 bit | 4.15 proxmox kernel | omvextrasorg 4.1.11
      omv-extras.org plugins source code and issue tracker - github

      Please read this before posting a question and this and this for docker questions.
      Please don't PM for support... Too many PMs!
    • His point of using raid 0 is still valid. If you lost one drive, everything would be gone. With unionfilesystems, you only lose data from the failed drive.
      omv 4.1.11 arrakis | 64 bit | 4.15 proxmox kernel | omvextrasorg 4.1.11
      omv-extras.org plugins source code and issue tracker - github

      Please read this before posting a question and this and this for docker questions.
      Please don't PM for support... Too many PMs!
    • iddqd wrote:

      elastic wrote:

      Ah. Going to jbod the discs. Would unionfilesystems be better? I've never heard of this before.
      If you are going jbod, I would look into stuff like SnapRAID to add some parity.
      Thank you, I will read up on it.

      Can anyone tell me if using a USB 3.0 SSD will work as a boot drive when connected to a USB 3.0 port?
      Case: U-NAS NSC-810
      Motherboard: ASRock - C236 WSI Mini ITX
      CPU: Core i7-6700
      Memory: 32GB Crucial DDR4-2133
    • ryecoaaron wrote:

      His point of using raid 0 is still valid. If you lost one drive, everything would be gone. With unionfilesystems, you only lose data from the failed drive.
      Could you explain how to do it? Actually, when I had experienced 1 hdd failure with unionfilesystem(mergerfs), I had no idea how to revive the lost data from the failed drive. So I returned back to omv 2.0 from 3.0.
    • canadavault wrote:

      when I had experienced 1 hdd failure with unionfilesystem(mergerfs), I had no idea how to revive the lost data from the failed drive
      You have to restore from backup or like Niemand said, use snapraid. The unionfilesystems aren't backup or redundancy.
      omv 4.1.11 arrakis | 64 bit | 4.15 proxmox kernel | omvextrasorg 4.1.11
      omv-extras.org plugins source code and issue tracker - github

      Please read this before posting a question and this and this for docker questions.
      Please don't PM for support... Too many PMs!
    • While OMV can simplify some of the setup you may want to read some tutorials to get an idea of what's going under the hood.

      github.com/trapexit/backup-and-recovery-howtos
      zackreed.me/setting-up-snapraid-on-ubuntu/
      zackreed.me/mergerfs-another-g…pool-your-snapraid-disks/

      Union filesystems tend to just literally create a union of several filesystems. That's it. RAID{1,5,6,Z} like redundancy requires another solution. SnapRaid is popular one. Some people even use a software RAID solution and create a few smaller arrays of drives and then union them so they can do more targeted upgrades of drives.

      Regardless of local redundancy you should always have a proper backup solution if the data is important. Some use CrashPlan. Some a cloud provider like Google GDrive or Amazon + some backup or mirroring software.

      Due to my years of data hoarding and my search to close random usability gaps I've found myself doing all of the above except the small RAID pools.
    • canadavault wrote:


      ryecoaaron wrote:

      His point of using raid 0 is still valid. If you lost one drive, everything would be gone. With unionfilesystems, you only lose data from the failed drive.
      Could you explain how to do it? Actually, when I had experienced 1 hdd failure with unionfilesystem(mergerfs), I had no idea how to revive the lost data from the failed drive. So I returned back to omv 2.0 from 3.0.

      As far as I understood it, the unionfs simply pool together different filesystems from different phyisical disk to act like they're one disk.
      If you take away one disk from the NAS the data in it will still be readable. So if one of the hdd fails, you will lose only those data while the others hdd and data will remain intact.
      Clearly, unless you have cold backups or some kind of parity, the data of the failed hdd is lost.
    • Ammiraglio wrote:

      As far as I understood it, the unionfs simply pool together different filesystems from different phyisical disk to act like they're one disk.
      If you take away one disk from the NAS the data in it will still be readable. So if one of the hdd fails, you will lose only those data while the others hdd and data will remain intact.
      Clearly, unless you have cold backups or some kind of parity, the data of the failed hdd is lost.
      You are correct.
      omv 4.1.11 arrakis | 64 bit | 4.15 proxmox kernel | omvextrasorg 4.1.11
      omv-extras.org plugins source code and issue tracker - github

      Please read this before posting a question and this and this for docker questions.
      Please don't PM for support... Too many PMs!