system is renaming docker interfaces which causes the network to stop wotking

  • Recently after replacing an HBA and booting my server running OMV5 I noticed that it won't connect to the network anymore. Not even the ethernet LEDs would light up. Upon checking dmesg I got the following:

    Code
    [ 27.013796] eth0: renamed from veth3acd072
    [ 27.123315] eth0: renamed from veth2c7a204
    [ 934.731916] veth3acd072: renamed from eth0
    [ 934.996108] veth7298456: renamed from eth1
    [ 1310.256532] eth0: renamed from vetha02b746

    I know what this problem is as I have experienced it on Raspberry Pi SBCs. It has to do with the system picking up the virtual interfaces created by Docker. The fix for it is normally easy:
    Just adding denyinterfaces veth* to /etc/dhcpcd.conf or editing /etc/NetworkManager/NetworkManager.conf to contain:

    Code
    [keyfile]
    unmanaged-devices=interface-name:veth*

    However, I can't do either of those since OMV uses a different networking system.

    I tried creating a rule in /etc/systemd/network/99-ignore-veth.network stating:

    Code
    [Match]
    Name=veth*
    
    [Link]
    Unmanaged=yes

    That did nothing.

    Can anyone guide in what I should do in order to replicate the above fix on OMV 5. Thanks.

    • Official Post

    OMV 5.x is EOL. You really should upgrade to 6.x


    I'm not sure why you are having this problem unless you still have some conflicting network packages installed. I have lots of OMV systems running docker and don't have this issue.


    What is the output of: dpkg -l | grep -iE "dhcp|network"

    omv 8.1.1-1 synchrony | 6.17 proxmox kernel

    plugins :: omvextrasorg 8.0.2 | kvm 8.0.7 | compose 8.1.5 | cterm 8.0 | borgbackup 8.1.7 | cputemp 8.0 | mergerfs 8.0 | scripts 8.0.1 | writecache 8.1.1


    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 am aware that the system is EOL and I should upgrade it, however it's a system at a remote location that I don't have physical access to and the way I interact through it is via a third party, so I put off updating it until I am there to diagnose any possible problems such as this. I will post the output of dpkg -l | grep -iE "dhcp|network" once said 3rd party will be free again to run the command (because I cannot access it remotely due to the completely dead networking)

    • Official Post

    Was the new HBA put in a different slot by any chance? I can see predictive naming changing the name of the network adapter if a different slot was used. But it shouldn't use existing docker virtual interface. I have four adapters in my system and see this:

    Code
    [    9.690362] e1000e 0000:00:19.0 enp0s25: renamed from eth2
    [    9.743771] igb 0000:05:00.0 enp5s0: renamed from eth0
    [    9.767563] sfc_siena 0000:03:00.1 ens1f1np1: renamed from eth3
    [    9.807392] sfc_siena 0000:03:00.0 ens1f0np0: renamed from eth1

    Did you install from the OMV iso?

    omv 8.1.1-1 synchrony | 6.17 proxmox kernel

    plugins :: omvextrasorg 8.0.2 | kvm 8.0.7 | compose 8.1.5 | cterm 8.0 | borgbackup 8.1.7 | cputemp 8.0 | mergerfs 8.0 | scripts 8.0.1 | writecache 8.1.1


    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!

    • Official Post

    the HBA was placed in the same slot where the old HBA was.

    Exact same HBA?

    omv 8.1.1-1 synchrony | 6.17 proxmox kernel

    plugins :: omvextrasorg 8.0.2 | kvm 8.0.7 | compose 8.1.5 | cterm 8.0 | borgbackup 8.1.7 | cputemp 8.0 | mergerfs 8.0 | scripts 8.0.1 | writecache 8.1.1


    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!

  • No, I exchanged an old unknown HBA (since that one periodically dropped sata devices as of recently) to an LSI Logic SAS3444E.

  • Below is the output of dpkg -l | grep -iE "dhcp|network" please excuse me omitting the architecture and the description since I had to copy this form an image of the screen.


  • Back to the original question, is it even possible to replicate the rules I asked about initially in omv or does the networking implementation not offer such capabilities?

  • chente

    Added the Label OMV 5.x
  • Running omv-firstaid and selecting the correct interface seems to have fixed the issue for long enough for me to upgrade to OMV 6.

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!