Building OMV automatically for a bunch of different ARM dev boards

  • I threw away the original Seagate 2.5" enclosure to somewhere.
    I will try to find it Monday.
    As per my note, its vid:pid is 0bc2:ab28.
    If I find it, I will post the test result on here.


    By the way, can you tell me more about the potential kernel fix?

  • My cousin sister found the enclosure in her room. :)
    The original Seagate enclosure worked fine with coherent_pool=1M option.
    I've flashed the latest OMV image cleanly too and it also worked nicely out of the box.

  • Hello everyone,


    I'm Frepke and new on the forum and waiting for my ordered Espressobin.
    I would like to run OMV on the Espressobin, is this possible already?
    And if yes, is this done trough the installer in Armbian or an already build image?


    Thanks for any help... ^^



    kr.,
    Frepke

    HP t630 Thin Cliënt (AMD Embedded G-Series GX-420GI | QuadCore | 8GB)
    7.0.4-2 (Sandworm) | 64 bit | pve-kernel-6.5 | omvextrasorg 7.0

  • I would like to run OMV on the Espressobin, is this possible already?

    I let a new OMV build run at the moment, will upload it later.


    Please keep in mind that this image is based on a WiP (work in progress) Armbian configuration using Marvell's legacy kernel with a few unresolved issues yet. So please consider this being a preview and not a fully stable/supported OMV image yet. At least I won't release any official OMV images for Espressobin unless mainline kernel support is there (since Marvell's 4.4 BSP kernel might not be supported that long)


    You can watch progress over there: https://forum.armbian.com/inde…velopment-efforts/&page=5 (please remember: it's WiP for a reason and currently only developer feedback is needed and no end-user support available)

  • I let a new OMV build run at the moment, will upload it later.
    Please keep in mind that this image is based on a WiP (work in progress) Armbian configuration using Marvell's legacy kernel with a few unresolved issues yet. So please consider this being a preview and not a fully stable/supported OMV image yet. At least I won't release any official OMV images for Espressobin unless mainline kernel support is there (since Marvell's 4.4 BSP kernel might not be supported that long)


    You can watch progress over there: https://forum.armbian.com/inde…velopment-efforts/&page=5 (please remember: it's WiP for a reason and currently only developer feedback is needed and no end-user support available)

    Thanks tkaiser,


    I hope there's a stable build available soon, but in the mean time I follow the Armbian forum.



    kr.,
    Frepke

    HP t630 Thin Cliënt (AMD Embedded G-Series GX-420GI | QuadCore | 8GB)
    7.0.4-2 (Sandworm) | 64 bit | pve-kernel-6.5 | omvextrasorg 7.0

  • I hope there's a stable build available soon, but in the mean time I follow the Armbian forum.

    Here you go: OMV_3_0_86_Espressobin_4.4.79.7z


    But I wouldn't call this stable atm -- quite the opposite to be honest. Starting with this release no SSH login as root is possible by default so you have to login through the web UI first, access the SSH service and allow root login. I already talked to @ryecoaaron about a mandatory password change on first login (that then also changes the root passwd) and still hope this gets implemented soon (re-enabling SSH login for root might then be an option after password change).


    I couldn't test any further with this image since It seems I'm somewhat unlucky with Espressobin so far (this is my 2nd board, I'm never able to access a writeable serial console, this 2nd board also overheats a lot). Won't look into this for the next three weeks since too busy with other stuff.

  • Here you go: OMV_3_0_86_Espressobin_4.4.79.7z
    But I wouldn't call this stable atm -- quite the opposite to be honest. Starting with this release no SSH login as root is possible by default so you have to login through the web UI first, access the SSH service and allow root login. I already talked to @ryecoaaron about a mandatory password change on first login (that then also changes the root passwd) and still hope this gets implemented soon (re-enabling SSH login for root might then be an option after password change).


    I couldn't test any further with this image since It seems I'm somewhat unlucky with Espressobin so far (this is my 2nd board, I'm never able to access a writeable serial console, this 2nd board also overheats a lot). Won't look into this for the next three weeks since too busy with other stuff.

    Thanks tkaiser,


    I thought I give it a try when my espressobin arrives, but I saw your message at the Armbian forum https://forum.armbian.com/inde…findComment&comment=36984. That made me doubt, maybe I'll wait for a more stable version of Armbian.


    kr.,
    Frepke

    HP t630 Thin Cliënt (AMD Embedded G-Series GX-420GI | QuadCore | 8GB)
    7.0.4-2 (Sandworm) | 64 bit | pve-kernel-6.5 | omvextrasorg 7.0

    Einmal editiert, zuletzt von Frepke ()

  • That made me doubt, maybe I'll wait for a more stable version of Armbian.

    Well, currently this looks like a simple hardware issue called overheating needing either hardware fixes (active cooling) or software (checking at which voltage the CPU is fed, if possible optimize this. We don't need to think about another countermeasure called throttling since currently I get these symptoms even with an idle board).


    Let's wait and see. I'll run the Linpack benchmark the next 12 hours with active fan and heatsink and if the board survives that IMO it's obvious that we've a hardware issue. See also http://espressobin.net/forums/topic/overheating/

  • I just wanna thank @tkaiser, Armbian and OMV for providing the best OMV installation since OMV 2.x. I just switched to it on Friday and I am more than pleased with my XU4 performance and overall behaviour. My USB 3.0 HDD (4TB Seagate Hub) does not have the UAS issues, luckily.

    Riddle me this, riddle me that
    Who is afraid of the big, black bat?
    I write on a blog (Romanian mostly)
    Testing (latest) OMV 6.x (HURRAY) on an Intel i5820K NAS (currently with proxmox kernel 6.2)

  • Here you go: OMV_3_0_86_Espressobin_4.4.79.7z


    But I wouldn't call this stable atm -- quite the opposite to be honest. Starting with this release no SSH login as root is possible by default so you have to login through the web UI first, access the SSH service and allow root login. I already talked to @ryecoaaron about a mandatory password change on first login (that then also changes the root passwd) and still hope this gets implemented soon (re-enabling SSH login for root might then be an option after password change).


    I couldn't test any further with this image since It seems I'm somewhat unlucky with Espressobin so far (this is my 2nd board, I'm never able to access a writeable serial console, this 2nd board also overheats a lot). Won't look into this for the next three weeks since too busy with other stuff.


    Hello tkaiser,


    A few days ago I received my espressobin from @Lupus. After I installed Putty on my MacBook, I set the environment variables and try to boot an image from the sd-card.


    With Armbian_5.32_Espressobin_Ubuntu_xenial_default_4.4.78.7z it's possible to boot. But I like to run OMV on the Espressobin, so I tried OMV_3_0_86_Espressobin_4.4.79.7z


    Unfortunately it's not possible for me to boot this image from an sd-card.


    Should it be possible to run this Image?



    Kr.,
    Frepke



    Verzonden vanaf mijn iPhone met Tapatalk

    HP t630 Thin Cliënt (AMD Embedded G-Series GX-420GI | QuadCore | 8GB)
    7.0.4-2 (Sandworm) | 64 bit | pve-kernel-6.5 | omvextrasorg 7.0

    Einmal editiert, zuletzt von Frepke ()

    • Offizieller Beitrag

    @tkaiser I know you aren't a big xu4 fan but what do you think of the odroid-hc1?

    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!

  • I know you aren't a big xu4 fan but what do you think of the odroid-hc1?


    Looks very promising! Some more info here.



    I already suggested few pages back to avoid USB3 cable/contact hassles by putting the USB-to-SATA bridge on the PCB with RK3328/ROCK64 in mind but Hardkernel already delivered. The idea to use the metal enclosure as giant heatsink is also great and they even thought about the 3.5" HDD use case, there's a 12V connector on the HC1 board that feeds the 12V pins on the SATA power connector. There are also some rumours wrt a HC2 (with the ability to connect two disks most probably using JMS567 as on the Cloudshell 2 which I won't consider a great solution. Adding either an USB hub or a port multiplier adds more complexity).


    Hardkernel sent out 2 HC1 review samples that should arrive soon and I'll report back here my findings (already curious how SSD performance will look like since on HC1 there's no USB3 hub in between host and disk). But to be honest: Now that XU4's Achilless heel is avoided (USB3 contact problems), that voltage-drops are less likely to happen (Cloudshell 1 problem) and that a good USB-to-SATA bridge is used avoding potential UAS hassles the only remaining question is about how efficient heat dissipation will be. HC1 single disk NAS performance will be as good as any x86 box maxing out Gigabit Ethernet anyway.

  • what do you think of the odroid-hc1?


    Small update: I'm still struggling with obscure network performance problems in my lab (only affecting XU4 and HC1 but other devices not for whatever reasons. Tried now the 3rd switch after my normal test switch died but to no avail). So still no full review and just some information.


    I collected some performance data and observations with regard to heat dissipation and thermal/throttling behaviour with OMV's kernel variant today: See comments here: https://www.cnx-software.com/2…-is-now-available-for-49/


    Also today over at Armbian we did some testing wrt AES performance. ODROID's Exynos SoC doesn't support ARMv8 crypto extensions (since being a 32-bit ARMv7 SoC with A15/A7 cores) so if we're talking about encryption/decryption performance (eg LUKS) other platforms might have a clear advantage here. Currently RK3328 on ROCK64 is best performer when judging by synthetic benchmarks (openssl --speed). ROCK64 AES single and multi threaded performance are 9 to 11 times better compared to HC1: https://forum.armbian.com/inde…findComment&comment=37829


    We're preparing a few more realistic tests covering LUKS encryption and then local IO performance (iozone tests) as well as NAS performance (LanTest). But this might take some time...

  • ODROID-HC1 mini review is online: https://forum.armbian.com/index.php?/topic/4983-odroid-hc1/


    Really impressive NAS performance especially with Ethernet IRQs moved from little to big core (which is something some users might want to revert to slightly lower overall consumption under load). But playing energy efficient NAS with HC1 is somewhat hard anyway (especially when compared with upcoming RK3328 boards with same performance but half the consumption -- but until there are RK3328 boards available that have also an onboard JMS578 or something similar HC1 is in a league of its own)

  • @ryecoaaron


    May I ask to replace all 'other arm images'? At Armbian we investigated about BTRFS (resizing, compression, performance, swap vs. zram) and here we go: http://kaiser-edv.de/tmp/NumpU8/


    Can you please exchange contents of https://sourceforge.net/projec…s/Other%20armhf%20images/ with the above?


    Changes: All images updated to kernel 4.12.9 and OMV 3.0.87, btrfs rootfs, zram instead of swap, countless tweaks/fixes. To document image integrity: The following is readme.txt contents:

    Currently Pine64/SoPine images will disappear but that's good since we're close to getting mainline kernel support for these devices.


    Other interesting devices (ROCK64, EspressoBin and others) will follow soon, same with some ODROID images (kernel updates applied, some u-boot finetuning needed).


    I made only tests with a couple of boards... but since all images are created fully automated from scratch it's only necessary to test one image per board family :)


    Banana Pi: http://sprunge.us/RYBQ
    Banana Pi Pro: http://sprunge.us/ONeZ
    PcDuino3 Nano: http://sprunge.us/WdaR
    Clearfog Pro: http://sprunge.us/SFcc

    • Offizieller Beitrag

    May I ask to replace all 'other arm images'?

    Sorry, been busy at VMworld in Vegas. I will start uploading them soon.

    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!