Unable to manage/activate network interface

I run RL9 and have a build-in network card (NC) on the motherboard (“eno1”) and an extra NC (“enp1s0”). The extra NC works just fine, but the built-in interface can not be managed. I tried to include/manage it with nmcli, but this does not work/change anything:

[root@server system-connections]# nmcli device status
DEVICE  TYPE      STATE      CONNECTION 
enp1s0  ethernet  connected  enp1s0     
eno1    ethernet  unmanaged  --         
lo      loopback  unmanaged  --         
[root@server system-connections]# nmcli device set eno1 managed yes
[root@server system-connections]# nmcli device status
DEVICE  TYPE      STATE      CONNECTION 
enp1s0  ethernet  connected  enp1s0     
eno1    ethernet  unmanaged  --         
lo      loopback  unmanaged  --         
[root@server system-connections]#
[root@server ~]# nmcli device show eno1
GENERAL.DEVICE:                         eno1
GENERAL.TYPE:                           ethernet
GENERAL.HWADDR:                         04:42:1A:AC:FE:2C
GENERAL.MTU:                            1500
GENERAL.STATE:                          10 (unmanaged)
GENERAL.CONNECTION:                     --
GENERAL.CON-PATH:                       --
WIRED-PROPERTIES.CARRIER:               off
IP4.GATEWAY:                            --
IP6.GATEWAY:                            --
[root@server ~]#

[root@server ~]# nmcli connection show
NAME    UUID                                  TYPE      DEVICE 
enp1s0  7a5fc8cd-1ab6-3d97-a9b2-ed4636b0ed7e  ethernet  enp1s0 
eno1    246dbe66-1b0a-4333-9e1a-01a57989fa98  ethernet  --     
eno1    8cfe34b4-46f0-48b8-872c-ad73c0c6ab6c  ethernet  --     
eno1    65fe63f7-dbb7-4637-8fb2-bfdecc7ce750  ethernet  --     
eno1    5b1fc545-4416-4b67-9821-5960df7d41fc  ethernet  --     
eno1    3098bd17-b182-4cf6-b890-3fc7f6990ffc  ethernet  --     
eno1    4c92c98b-73c5-4fbc-9a23-f5aed87206d7  ethernet  --     
[root@server ~]# 

Can please anyone help me to point to what is wrong here? The card is apparently identified, but remain unmanageable, but there seem to be many UUID for the same card???. Thanks for any hint to a solution!
/Ralf

What is the device ID of that interface? lspci -nn | grep -i net should show the ID.

Thanks jlehtone for the quick reply!!

I get:

[root@server ~]# lspci -nn | grep -i net
00:1f.6 Ethernet controller [0200]: Intel Corporation Ethernet Connection (14) I219-V [8086:15fa] (rev 11)
01:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8161] (rev 15)
[root@server ~]#

How do I use that?

I also tried to use: nm-connection-editor, but when storing the info, I get:

** Message: 17:06:26.547: Connection validates and can be saved

(nm-connection-editor:20827): dconf-WARNING **: 17:06:34.183: failed to commit changes to dconf: Failed to execute child process “dbus-launch” (No such file or directory)

And no difference. The eno1 interface is still “unmanaged”:

[root@server ~]# systemctl restart NetworkManager.service
[root@server ~]# systemctl status NetworkManager
● NetworkManager.service - Network Manager
     Loaded: loaded (/usr/lib/systemd/system/NetworkManager.service; enabled; vendor preset: enabled)
     Active: active (running) since Tue 2022-10-04 17:11:31 CEST; 8s ago
       Docs: man:NetworkManager(8)
   Main PID: 22786 (NetworkManager)
      Tasks: 4 (limit: 98144)
     Memory: 3.4M
        CPU: 44ms
     CGroup: /system.slice/NetworkManager.service
             └─22786 /usr/sbin/NetworkManager --no-daemon

Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1131] device (enp1s0): state change: ip-config -> ip-check (r>
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1141] device (enp1s0): state change: ip-check -> secondaries >
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1142] device (enp1s0): state change: secondaries -> activated>
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1144] manager: NetworkManager state is now CONNECTED_LOCAL
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1145] manager: NetworkManager state is now CONNECTED_SITE
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1146] policy: set 'enp1s0' (enp1s0) as default for IPv4 routi>
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1180] device (enp1s0): Activation: successful, device activat>
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1183] manager: NetworkManager state is now CONNECTED_GLOBAL
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.1185] manager: startup complete
Oct 04 17:11:32 server.hartings.se NetworkManager[22786]: <info>  [1664896292.2538] agent-manager: agent[fd26cf75eb524b0d,:1.77/org.gnome.S>
[root@server ~]# nmcli device status
DEVICE  TYPE      STATE      CONNECTION 
enp1s0  ethernet  connected  enp1s0     
eno1    ethernet  unmanaged  --         
lo      loopback  unmanaged  --         
[root@server ~]# 

So I am out of ideas right now. Any help is appreciated!

Usually, if there is no driver for device – Red Hat has not added for new or has omitted old driver – one goes to ELRepo ELRepo | DeviceIDs to see whether they have packaged one.

Alas, the e1000e kernel module in RL9 appears to support this device:

$ modprobe -c | grep 15FA
alias pci:v00008086d000015FAsv*sd*bc*sc*i* e1000e
$ modinfo e1000e | grep 15FA
alias:          pci:v00008086d000015FAsv*sd*bc*sc*i*

The driver is relatively new though, like Linux Kernel Driver DataBase: CONFIG_E1000E: Intel(R) PRO/1000 PCI-Express Gigabit Ethernet support lists:

lkddb pci 8086 15fa : CONFIG_E1000E # in 5.5–5.19, 6.0, 6.0+HEAD

Red Hat’s bugzilla does show at least one bug for it: https://bugzilla.redhat.com/buglist.cgi?quicksearch=e1000e
That bug (2094200) has different symptoms.

Can you test with some different OS&driver, e.g. Live Fedora/Ubuntu USB? If they do work, then hardware should be ok.

Thanks for this in depth reply!
I checked the links you provided and realize there is an issue with this interface. It’s a relatively new motherboard , but not that new, 2 years old. And from Intel, not a goofy third party supplier. This really surprises me.
I think a quicker solution is perhaps buy a new PCI express NC rather then doing several tests. I had never expected this kind of issue. I need to get this machine operational as soon as possible!
Again, many thanks for this great help/info! I am glad I chose RL9 instead of Alma Linux. The forum over there is not as quick and knowledgeable as here!

Could this onboard device be possibly disabled in the bios?

Thanks jbkt23 for the suggestion. As the interface is detected in the OS, it should be active in the BIOS I guess, but I will check just to be sure.
The issue with the driver is there anyhow, so a new NC is probably a better and quicker solution. They are relatively cheap and easy to buy, so no problem.

One more thing worth trying in case the module basically doesn’t support your particular card, would be to use ELREPO and kernel-ml which is 6.x kernel.

[root@rocky9 ~]# dnf info kernel-ml --enablerepo=elrepo-kernel
Last metadata expiration check: 0:00:24 ago on Wed 05 Oct 2022 09:59:39 AM CEST.
Available Packages
Name         : kernel-ml
Version      : 6.0.0
Release      : 1.el9.elrepo
Architecture : x86_64
Size         : 11 k
Source       : kernel-ml-6.0.0-1.el9.elrepo.src.rpm
Repository   : elrepo-kernel
Summary      : The Linux kernel. (The core of any Linux kernel based operating system.)
URL          : https://www.kernel.org/
License      : GPLv2 and Redistributable, no modification permitted.
Description  : The kernel-ml meta package.

Obviously you can use the dnf install command on that info above, to install and check it out. One way of ruling out any potential issues with incompatible modules.

Thanks iwalker! Yes, this could be an alternative. Just don’t want to mess up the installation, as I am a bit in a hurry to get this server up and running as my on-line active server. It’s still on-the-side right now, with an older server still doing the job.
Are there any known conflicts with this elrepo kernel and the rest of the standard RL9 installation? Are other RL9 standard packages known to be compatible with this new kernel? Has this new kernel been tested or should it be regarded as an “experimental” kernel? I prefer a reliable and stable server, which doesn’t require fixing…
If any issues are showing up during operation, can I just re-install the official RL9 kernel package again and that will remove the elrepo kernel and install the standard RL9 kernel?
I hope you don’t mind me asking; I am just carefull, as I put quite some time in configuring this server :slight_smile:

Hi, according to:

[root@rocky9 ~]# dnf install kernel-ml --enablerepo=elrepo-kernel
Last metadata expiration check: 1:35:19 ago on Wed 05 Oct 2022 09:59:39 AM CEST.
Dependencies resolved.
==================================================================================================================================
 Package                           Architecture           Version                             Repository                     Size
==================================================================================================================================
Installing:
 kernel-ml                         x86_64                 6.0.0-1.el9.elrepo                  elrepo-kernel                  11 k
Installing dependencies:
 kernel-ml-core                    x86_64                 6.0.0-1.el9.elrepo                  elrepo-kernel                  56 M
 kernel-ml-modules                 x86_64                 6.0.0-1.el9.elrepo                  elrepo-kernel                  51 M

Transaction Summary
==================================================================================================================================
Install  3 Packages

Total download size: 107 M
Installed size: 148 M
Is this ok [y/N]: 

as you can see, it’s just installing the relevant kernel packages, so no dependencies outside of this that could cause you problems.

As for the kernel:

[root@rocky9 ~]# dnf repolist --all | grep -i elrepo
elrepo                  ELRepo.org Community Enterprise Linux Repositor enabled
elrepo-extras           ELRepo.org Community Enterprise Linux Extras Re disabled
elrepo-kernel           ELRepo.org Community Enterprise Linux Kernel Re disabled
elrepo-testing          ELRepo.org Community Enterprise Linux Testing R disabled

as you can see from the previous command that it comes from elrepo-kernel rather than elrepo-testing, would mean it’s pretty much as stable as you are going to get for any kernel you are using. That said, if you install and don’t like it, you can choose to boot back to the previous kernel as it will still be on the list, and you can remove the newer kernel at any point without any issues. Kernels are only removed if, 1: they have exceeded the limit of 3 or 5 or whatever the default is, or 2: you uninstalled it manually yourself. Generally otherwise, the latest 3 or 5 will be on your system.

2 Likes

Thanks again iwalker!
Yes, that is certainly reassuring! I will take a look at it.

The kernel-ml packages are maintained by ELRepo. They build from sources of “mainline” kernel. If a bug is found in mainline kernel, then kernel developers patch it and ELRepo rebuilds their package.

The kernel package in Rocky is built by Rocky team from released sources of RHEL. Red Hat patches those sources and releases built package for RHEL (and sources), when they deem it necessary.

AFAIK, Linux kernel development prefers not to break compatibility and therefore mainline kernel should be compatible even with EL platform.

1 Like

Thanks jletone! That is indeed good to know.
I will try the kernel-ml and let you know if it solved my problem.

Update: I installed the elrepo kernel-ml but this makes no difference:

[root@server ~]# uname -r
6.0.0-1.el9.elrepo.x86_64
[root@server ~]# nmcli device status
DEVICE  TYPE      STATE      CONNECTION 
enp1s0  ethernet  connected  enp1s0     
eno1    ethernet  unmanaged  --         
lo      loopback  unmanaged  --         
[root@server ~]# nmcli device set eno1 managed yes
[root@server ~]# nmcli con mod eno1 ipv4.gateway 192.168.1.254
[root@server ~]# nmcli con mod eno1 ipv4.address 192.168.1.94
[root@server ~]# nmcli device status
DEVICE  TYPE      STATE      CONNECTION 
enp1s0  ethernet  connected  enp1s0     
eno1    ethernet  unmanaged  --         
lo      loopback  unmanaged  --         
[root@server ~]# nmcli con up eno1
Error: Connection activation failed: No suitable device found for this connection (device enp1s0 not available because profile is not compatible with device (mismatching interface name)).
[root@server ~]# 

So, either the 6.0 kernel is not compatible with the on-board NC either, or the interface is defect (hardware issue). In any case, I cannot get it to work. I will try with a separate, new PCI express NC card tomorrow.
In the mean time I have removed the kernel-ml packages, just to run a clean Rocky 9 distribution :slight_smile:

Update: I installed a separate PCI Express NC and this worked just fine! For now I leave the non-working built-in NC on the motherboard. I needed two NC’s and I have now two.

Thanks to all who have bothered to read my questions and give valuable answers. I learned a lot again!
That’s why we have such forums! They are as important as the software they are discussing!

1 Like