Server starts on its own after shutdown (Asus H87I-Plus)

  • Hi,
    i've got a problem with my server. :)
    For a while everything worked fine. But the last few days i recognized a strange behavior. Everytime the server starts immediately (~2-3sec (HDDs spun down) after a normal shutdown. There is no difference in operation between a manual (over WebUI or cli: shutdown -h now and an automatic (plugin "Autoshutdown") shutdown. My server runs on OMV 1.0.23 (standalone release) and I also flashed the newest Bios version (2002).


    My Log ID is F4laAiSH


    Edit:
    I searched a little bit across the web and it seems to be a hardware problem of my motherboard (Asus H87I-Plus (C2)). :(


    Here are some links i found about the issue:

    My linux skills are unfortunately to low to solve the problem on my own if at all possible. I just found the option "WOL" in the network settings of OMV and activated it. Don't know what exactly it does because waking up the server over network works fine before. Hopefully it will solve the spontaneous "power on" issue. I will keep an eye on it the next days.


    Edit 2:
    Ok, the WoL check mark over the network settings of OMV WebUI has no effect on the starts by itself.


    Edit 3:
    Tested al lot the last few weeks. Swap RAM and PSU, only the OS HDD is connected, unplug the network cable, backports kernel with newest intel network driver (3.1.0.2). Nothing changed the issue. Only the deactivation of WoL in the Bios solved the problem, as described in the first link I posted. Meanwhile I bought a new board (ASRock Z87E-ITX) and everything performs flawlessly. Can't work without WoL :)

    • Offizieller Beitrag

    I've had this problem forever. In my case, it would do exactly as you describe.. instead of shutting down, it would power down, then boot back up about 2-4sec later. After that, if I issued the shutdown -h now command again.. the machine would shutdown normally.


    How does lspci identify your device? (.. lowercase L)
    03:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B PCI Express Gigabit Ethernet controller (rev 09)


    I've also always noticed this message in my logs, but other than this shutdown quirk.. the device seemed to work reasonably well:
    W: Possible missing firmware /lib/firmware/rtl_nic/rtl8105e-1.fw for module r8169


    Well, I did some Googling yesterday and came across this thread. https://bbs.archlinux.org/viewtopic.php?id=173648&p=1 . Post #2 has a link to what is supposed to be the proper firmware.


    Before installing the driver, I had the backport kernel installed. After that, I downloaded the new driver onto my OMV box and installed it, and it's still doing it. This problem (for me) seems isolated to WOL. If I use WOL, guaranteed I will have to do the "shutdown" twice. If I don't use WOL, and instead manually power on the NAS.. usually the first shutdown works correctly.


    I've come to the conclusion that realtek Linux support just absolutely SUCKS... I'll probably buy an Intel card this week because this issue has finally just got to the point where, for some reason.. it's really irritating me.. :). Who knows, I might even finally order an SSD to.

  • Thanks for the replies so far!


    @CaNsA
    I set this point to off before i installed OMV. So it should not be the issue. Thanks anyway!


    @KM0201
    I'm not so familiar with the comandline. Do i have to navigate to a specific folder and execute "lspci |" there? Furthermore you talking about a Realtek nic, if i'm not completely wrong, but my board has an Intel nic (Intel I217-V) onboard.

    • Offizieller Beitrag

    Hmm... I saw conflicting info.. a thread on another forum I read earlier, said it was a realtek.. thus why I thought our probs might be related, or someone may have misposted as well...


    Edit: But to answer your question... no.. lspci can be executed anywhere.

    • Offizieller Beitrag

    I have the same problem with my system, is an ECS H61H2-I3.


    A cron poweroff command sometimes just reboots, sometimes shutsdown. My solution for now is to put a second cron job with a poweroff command 5 minutes later after the first one.


    I am gonna try disabling that second command, maybe with the new kernel or backports it works better.

  • A few days have passed but the problem is still there :)
    Unfortunately the backports kernel does not resolve the issue. My last hope is to install the newest nic-driver (3.1.0.2) from intels website, because the backports 3.16 runs with a 19 month old driver version (2.3.2). I downloaded the driver as a tar.gz file but I`m unable to create a installation file. It seems, that the "make" package is missing in omv. After extracting the tar.gz file and navigating to the .../src subfolder in the new driver folder, I tried the command:


    make install


    But I only get the following error message:


    # -bash: make: Kommando nicht gefunden.


    How can I install the newest driver or the "make" package? I'm not very familiar with linux at all but I will try my best.

  • Hi,
    thanks for the fast reply! Ok, I will try to install both parts. I've found the README at my first installation attempt but failed at the "make" command, because of my lack of knowledge :D


    Edit: Ok, the make command works but i get the error message:
    Makefile:108: *** Linux kernel source not configured - missing autoconf.h. Schluss


    @subzero79 Is there anything I need to keep in mind once I start compiling the driver modules?

  • Hi there.


    I have done a brand new build an install on a Asrock Z97E-ITX/ac.
    Since the hardware is so new - i felt it best to use the backports kernel.
    i could not get the intel drivers to install after upgrading to backports because of this error... missing autoconf.h - i wasted a good 3 to 4 hours trying all sorts of things.


    So - i re-installed - started over.


    Firstly the e1000e drivers included in the 3.2 kernel are old - and there are many reports of poor nic performance - solved by installing the latest intel drivers manually.


    i can confirm that there are no errors when building and installing new e1000e drivers on the default 3.2 kernel. installed, tested and working just fine. i did this straight after installing from scratch and after running omv-firstaid.


    i then installed the backports, rebooted, and tried building the driver for the new backports 3.16 kernel.
    "make install" does not work because "*** Linux kernel source not configured - missing autoconf.h."


    Now i am really just playing around with this new install and i am happy to test anything and everything - it takes about 10 minutes to re-install if i mess it up.


    I would like to try upgrade to 3.16 backports manually (not using the plugin). that might tell us something about whether or not something is not 100% with the plugin.


    i am not that sharp though - a little rusty - but i am willing to give it a go.

    • Offizieller Beitrag

    Why do you need to compile drivers for the 3.16 kernel?


    As for your error, the plugin isn't the problem. The kernel source package isn't installed.

    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!

  • e1000e: Intel(R) PRO/1000 Network Driver - 2.3.2-k (from dmesg |grep e1000e)


    just rebooted to default kernel.... i succeeded in updated the intel driver on that:-


    e1000e: Intel(R) PRO/1000 Network Driver - 3.1.0.2-NAPI

    • Offizieller Beitrag

    For other people asking this same question, how did you fix the autoconf error?

    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!

  • I have not been able to solve the error. I can build the driver on the default 3.2 kernel and install it successfully... But I cannot build the driver on the 3.16 kernel. Hence the driver on the 3.16 kernel is still the old Intel driver.


    I am in a pickle now... Would like the most recent kernel because of his new the hardware is... But it is more important to have the newer more stable Intel drivers for the LAN... After all the LAN performance is integral to the OMV experience.

  • I see my response above isn't very clear -


    Basically - the default install on the 3.2 kernel allows me to update the intel LAN drivers to the most recent available.


    After doing the update on the intel drivers on the 3.2 kernel - i updated the omv install to backports kernel 3.16. This cannot be updated to the newest intel drivers as you pointed out the kernel source package isn't installed.


    So - snookered - i am a beginner - not knowledgeable to solve that issue.


    i have time to play with the system - i need to deploy it as a zoneminder and plesk server in about 2 weeks - then i need to be able to forget about it - once i know it is stable.

    • Offizieller Beitrag

    I figured it out :) This script will download the driver, compile, and install it.


    As root:


    Code
    wget http://omv-extras.org/intel.sh
    chmod +x intel.sh
    ./intel.sh

    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!

Jetzt mitmachen!

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