After restart, can't load web GUI

  • After I restart I can't ever pull up the web interface. My server is headless, so have had to attach a monitor, log in to the CLI and run 'dhclient'.
    Before that command "ip addr" shows the default local IP.
    As soon as I manually run this dhclient command, the webgui loads instantly.
    This wasn't the case until I moved the server into another room, attached to another switch (layer 2, un-managed). At that time, I also removed a graphics card and sound card. I've got to assume those are totally unrelated.


    This seems weird. Any ideas?

    • Offizieller Beitrag

    I've got to assume those are totally unrelated.

    I agree, it's probably coincidence, but strange things happen.


    The only thing I could come up with would be ESD damage while removing the cards. But that would be a long shot in that it's hard for hands and arms not to be grounded by the edges of a metal case when working inside.
    The move to another room and another switch shouldn't have anything to do with it. Removing the video and sound cards might change something in the PCI bus, but I can't see how that would affect the NIC (or configuration) in the very specific manner you're describing.


    Out of curiosity, have you reinstalled the video and sound card to see if there's any effect?
    (And)
    Have you tried statically addressing the server, bypassing DHCP altogether? Pick an address, outside of your routers DHCP scope, and give it a try.

  • Please have a look at systemctl status.
    Which service does manager your network? NetworkManager? Networkd?
    Please provide output of the corresponding config files.
    It might for example be in /etc/systemd/network.

  • The only thing I could come up with would be ESD damage while removing the cards.


    Out of curiosity, have you reinstalled the video and sound card to see if there's any effect?
    (And)
    Have you tried statically addressing the server, bypassing DHCP altogether? Pick an address, outside of your routers DHCP scope, and give it a try.

    I wondered about that too. Although this would be very specific damage, since the NIC will grab it's IP when manually instructed to. And after that functions perfectly.


    I did re-install, but it had no effect.


    I forgot to mention that from the start I've had the server under a DHCP reservation. I could try a static assignment on the server directly, but would really prefer to fix the issue properly.


    Please have a look at systemctl status.
    Which service does manager your network? NetworkManager? Networkd?
    Please provide output of the corresponding config files.
    It might for example be in /etc/systemd/network.

    I ran that command but couldn't see anything of interest. What should I be looking for?
    I found no references to network in that output.
    Where do I see which service manages my network?


    Also I couldn't find any config files in the directory you mentioned. Where else should I look?


    I did notice that systemctl status showed the server as degraded. Looked into that further and it said the disk quota system had failed. Not sure why, or how to fix that.

  • "systemctl --state=failed" didn't show anything.

    Zitat

    0 loaded units listed. Pass --all to see loaded but inactive units, too.


    To show all installed unit files use 'systemctl list-unit-files'.


    Now after a restart "systemctl status" shows 2 failed units.
    Running "systemctl -all" shows me that "quotaon.service" has failed again, and now "networking.service" has failed.
    I should mention that the last time I checked the status networking was active and running. Only quotaon was down.



    Also when I was trying to enable WoL in the ethernet interface options, I got this error:

    Zitat

    Error #0:OMV\ExecException: Failed to execute command 'export PATH=/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/local/sbin; export LANG=C.UTF-8; systemctl start 'networking' 2>&1' with exit code '1': Job for networking.service failed because the control process exited with error code.See "systemctl status networking.service" and "journalctl -xe" for details. in /usr/share/php/openmediavault/system/process.inc:182Stack trace:#0 /usr/share/php/openmediavault/system/systemctl.inc(86): OMV\System\Process->execute(Array, 1)#1 /usr/share/php/openmediavault/system/systemctl.inc(146): OMV\System\SystemCtl->exec('start', NULL, false)#2 /usr/share/openmediavault/engined/module/networking.inc(44): OMV\System\SystemCtl->start()#3 /usr/share/openmediavault/engined/rpc/config.inc(194): OMVModuleNetworking->startService()#4 [internal function]: OMVRpcServiceConfig->applyChanges(Array, Array)#5 /usr/share/php/openmediavault/rpc/serviceabstract.inc(123): call_user_func_array(Array, Array)#6 /usr/share/php/openmediavault/rpc/serviceabstract.inc(149): OMV\Rpc\ServiceAbstract->callMethod('applyChanges', Array, Array)#7 /usr/share/php/openmediavault/rpc/serviceabstract.inc(565): OMV\Rpc\ServiceAbstract->OMV\Rpc\{closure}('/tmp/bgstatusTb...', '/tmp/bgoutputn7...')#8 /usr/share/php/openmediavault/rpc/serviceabstract.inc(159): OMV\Rpc\ServiceAbstract->execBgProc(Object(Closure))#9 /usr/share/openmediavault/engined/rpc/config.inc(213): OMV\Rpc\ServiceAbstract->callMethodBg('applyChanges', Array, Array)#10 [internal function]: OMVRpcServiceConfig->applyChangesBg(Array, Array)#11 /usr/share/php/openmediavault/rpc/serviceabstract.inc(123): call_user_func_array(Array, Array)#12 /usr/share/php/openmediavault/rpc/rpc.inc(86): OMV\Rpc\ServiceAbstract->callMethod('applyChangesBg', Array, Array)#13 /usr/sbin/omv-engined(536): OMV\Rpc\Rpc::call('Config', 'applyChangesBg', Array, Array, 1)#14 {main}


    Reading through that I noticed the bit about "See "systemctl status networking.service"". Here is the output.


    I can send someone the journalctl -xe output, but it's pages and pages long, so didn't want to post it here.

  • Please provide ip link show.
    Are there multiple ethernet ports? Did you set one to be forced up which is not connected anymore?
    /sbin/ifup -a --read-environment [b](code=exited, status=1/FAILURE)[/b]
    I think we should solve this one. Are there any additional logs? What happens if you restart that service after boot completed?

  • Here's the ip link show output.

    There's just the one onboard NIC, and I haven't made any changes to networking settings.


    What other logs would you like? Where do I find them?


    It's weird that enp2s0 keeps being referenced, but isn't in the list. It seems like @crashtest might have been correct about the system's config having changed because of the video and sound card removal.

  • Ok, did you put in or remove pcie devices?
    The nic it tries to configure is located in another (as it is an onboard device virtual) pcie slot than the one existing. This happens sometimes with consumer hardware. This may be caused by adding or removing devices as well as changing bios settings.
    This can be fixed quite easily. Just reconfigure the network settings and set the old settings for enp1s0 instead of enp2s0.
    Edit: just read you did in fact remove pcie cards. Some consumer hsrdware unfortunally creates virtual pcie slots for onboard hardware behind the last slot used. This is hardware specific behaviour which can lead to the problems encountered.

  • Ok, did you put in or remove pcie devices?

    Yes! A pcie video card, and a pcie sound card.
    I did add some new hard drives, but other than that I didn't add or change anything else.


    Ah, that is interesting. I don't understand why that's the behaviour. It seems like it would cause a lot of needless issues. And yes, it's an ASUS consumer mainboard from a number of years ago.


    Under System > Network > Interfaces, I can see enp2s0 listed, but I can't edit it.
    Under Add, I can start the creation process for a new ethernet device, and enp1s0 is listed, but when I click save, I get the following error.

  • This is hardware specific behaviour which can lead to the problems encountered

    Only with inflexible network config mechanisms like Debian's ifupdown -- with Network Manager for example it doesn't matter whether the Ethernet device is called enp2s0 or enp1s0 or eth0 or enx0dea1db2e3af or whatever as long as there's at least one Ethernet interface.


    @sonofwatt problem is that Debian's implementation of predictable network interface names fails in situations like yours where removing/adding PCIe cards changes PCIe slot numbers for other components (so it seems the network card moved now to a different slot). The idea of these predictable interface names is good and definitely an improvement over the former eth schema where strange stuff with more than one NIC could happen all the time. But the concept can not deal that good with a situation where a card logically 'moved'. As almost everything in 'stock Debian' it's made for servers only.


    Maybe it would've been the easiest option to simply replace the old interface name with the new one: sed -i 's/enp1s0/enp2s0/g' /etc/network/interfaces (followed by a reboot or systemctl restart networking). But if the interfaces file is now already messed up it's too late.


    @votdev: when looking through your OMV5 changes I understood that primary interface will still be handled by ifupdown? And networkd also relies on the predictable interface names so no improvement with situations like this?

  • I am not all too familiar with the omv specific commands, but maybe omv-firstaid can reconfigure the network correctly.
    I personally would go with the sed mentioned above, than grep for enp1s0 and remove all doubling entries.
    @tkaiser I agree that NetworkManager has a more robust behaviour in this situation.

    Einmal editiert, zuletzt von tkaiser () aus folgendem Grund: omv misspelling

  • Ran sed -i 's/enp1s0/enp2s0/g' /etc/network/interfaces which I think worked.


    Then ran /etc/init.d/networking restart but it gave me the following error.

    Zitat

    [....] Restarting networking (via systemctl): networking.serviceJob for networking.service failed because the control process exited with error code.
    See "systemctl status networking.service" and "journalctl -xe" for details.
    failed!


    Carried on with omv-firstaid after the error. I selected option 1 (Configure network interface) on the omv-firstaid menu list. Noticed that the interface name had updated. Yes to config IPv4, Yes to DHCPv4, No to IPv6, Yes to WOL.
    Quick restart and the GUI came right up after a few moments!


    Thanks guys!

Jetzt mitmachen!

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