Has anyone tried using none 24/7 hr disks on hardware raid ?

  • I'm toying with the idea of trying this using 2 .1/2" fairly low powered disks and wondered if any one has done this before. The disks might best be described as high endurance ruggedised lap top drives. They are more usually used for robotics and in the automotive field.


    All drives will be the same and driven via a mini sas sata raid controller.


    Personally I can't see how there can be a problem providing the controller obeys the signals from the drives and also believe that some people have used ordinary laptop disks in server raid arrays.


    John
    -

    • Offizieller Beitrag

    Doesn't sound like they would be fast but I can't think of any reason they wouldn't work. I've used other industrial components with OMV before and not had problems.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


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


    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!

  • Double-stitched is always better :rolleyes:

    OMV 3.0.100 (Gray style)

    ASRock Rack C2550D4I C0-stepping - 16GB ECC - 6x WD RED 3TB (ZFS 2x3 Striped RaidZ1) - Fractal Design Node 304 -

    3x WD80EMAZ Snapraid / MergerFS-pool via eSATA - 4-Bay ICYCube MB561U3S-4S with fan-mod

  • Personally I can't see how there can be a problem providing the controller obeys the signals from the drives and also believe that some people have used ordinary laptop disks in server raid arrays.

    Please look at your other thread what I have written.


    In your headline you were talking about hardware raid. For hardware raid which has typically a dedicated raid controller you need drives which support TLER. But you are right, that a lot of people have used desktop drives in server raid arrays. But then they use some kind of software raid like mdadm or ZFS.

    OMV 3.0.100 (Gray style)

    ASRock Rack C2550D4I C0-stepping - 16GB ECC - 6x WD RED 3TB (ZFS 2x3 Striped RaidZ1) - Fractal Design Node 304 -

    3x WD80EMAZ Snapraid / MergerFS-pool via eSATA - 4-Bay ICYCube MB561U3S-4S with fan-mod

  • Actually I meant laptop drives on hardware raid controllers. Cheap skate approach. What I don't know is how well it works out. I just know that it has been done. Sort oh well disks fail anyway.


    Now I know what the drives need I found a write up of why


    https://forums.freenas.org/ind…support-on-a-drive.27126/


    There is a bit more here and an eg of a laptop drive that supports it


    https://raid.wiki.kernel.org/i…,_and_what_is_a_device%3F


    Disk like most things have evolved over time and that makes me wonder if this area isn't as straight forwards as it was. The storage density on disks has gone way up and I have seen some evidence that drives regularly correct errors via parity corrections. Only one drive but this particular one reports parity corrections separately. It has zero read errors but a lot of parity corrections. This action must have some sort of time out feature. Maybe the time can be set in the controllers - sounds like it should be. My chosen used raid controller hasn't arrived yet so will have to look around for data on it.


    Finding manuals on lsi controllers is often difficult. It's this one


    LSI00202 LSI Megaraid 9260-8i SAS/SATA 6GB/s RAID controller


    I can plug in one of the drives I want to use later today and see if the feature is supported. Also look and see if mdadm has a similar feature.



    John


    -

  • LSI00202 LSI Megaraid 9260-8i SAS/SATA 6GB/s RAID controller

    Some people use this kind of controllers which a special kind of IT firmware as pure HBA (Host Bus Adaptor) which disables the hardware raid features (I think) but provide the necessary sata or sas ports. It´s more or less a sata port extension. The raid functionality is then done in software and controlled by the OS.
    Personally I have no experience with this kind of HBAs.


    Some more information here in this quite long ZFS thread: Instal ZFS-Plugin & use ZFS on OMV and some other threads here.

    OMV 3.0.100 (Gray style)

    ASRock Rack C2550D4I C0-stepping - 16GB ECC - 6x WD RED 3TB (ZFS 2x3 Striped RaidZ1) - Fractal Design Node 304 -

    3x WD80EMAZ Snapraid / MergerFS-pool via eSATA - 4-Bay ICYCube MB561U3S-4S with fan-mod

    • Offizieller Beitrag

    I have a 9260. It doesn't have IT firmware though. It works fine with OMV otherwise.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.4 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


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


    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 drives do support it - sct. Smartctl was turned off initially and it took a while to realise that I needed the -d sat option as the drive was in a usb dock. I've found that there is no need to mount drives the usual way for this sort of thing. Not sure how but the system mounts them and when removed spacefm shows them as empty mounts, also allows those to be deleted. I've also formatted drives mounted this way but when that's done it can be really bad idea to use kde's eject button afterwards. If that's clicked before using open with file manager etc it functions as mount for something else which can be confusing.


    Much easier to find data on the raid controller since the last time I used one from LSI. May as well post a link as there are a number of these about.


    https://www.broadcom.com/suppo…s&pn=MegaRAID+SAS+9260-8i


    || Lots of reading to do but the card looks promising. It can even support drive spin down for power saving and also hot fix if the drives support hot plug. It looks like an easy way to extend the number of drives that can be fitted - electrically anyway. Mechanically may be a bit of a challenge. I'm not so sure about 2x 2 1/2 drive to 3 1/2 converters due to cooling.


    John
    -

Jetzt mitmachen!

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