Automatic modprobe (enable Linux Kernel module) at boot time

Written by Claudio Kuenzler - 0 comments

Published on February 14th 2023 - Listed in Linux Kubernetes Systemd

While fiddling with kubeadm to run a local Kubernetes cluster, I ran into an error:

root@kube1:~# kubeadm init --config=kubeadm-config.yaml --upload-certs
[init] Using Kubernetes version: v1.24.1
[preflight] Running pre-flight checks
error execution phase preflight: [preflight] Some fatal errors occurred:
    [ERROR FileContent--proc-sys-net-bridge-bridge-nf-call-iptables]: /proc/sys/net/bridge/bridge-nf-call-iptables does not exist
[preflight] If you know what you are doing, you can make a check non-fatal with `--ignore-preflight-errors=...`
To see the stack trace of this error execute with --v=5 or higher

This error showed up although the sysctl config contains the relevant values:

root@kube1:~# cat /etc/sysctl.d/kubernetes.conf
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
net.ipv4.ip_forward = 1

But when manually looking inside the procfs, the "bridge" part is missing in the path:

root@kube1:~# ll /proc/sys/net/
core/      ipv4/      ipv6/      netfilter/ unix/

A "bridge" directory should be there so sysctl can actually set these values.

In this situation, the "bridge" directory is missing because the relevant Kernel module (br_netfilter) is not loaded:

root@kube1:~# lsmod|grep netfilt
[... nothing ...]

Of course you can now run modprobe to manually load the module:

root@kube1:~# modprobe br_netfilter
root@kube1:~# lsmod|grep netfilter
br_netfilter           28672  0
bridge                176128  1 br_netfilter

But this won't survive a reboot.

As a workaround you could place the command "modprobe br_netfilter" into /etc/rc.local, once execution of rc.local is enabled. But there's a proper solution on distributions with Systemd: /etc/modules-load.d or /etc/modules.

By simply adding the module name(s) to load upon boot time into one of the *.conf files inside /etc/modules-load.d, these modules will be loaded during boot. As /etc/modules-load.d/modules.conf is a symlink to /etc/modules, I'm using this file:

root@kube1:~# echo "br_netfilter" >> /etc/modules

Of course you could also use a dedicated .conf file for this purpose.

After a system reboot, you will notice that br_netfilter is now listed under lsmod.

The Kubernetes cluster can now be created with kubeadm:

root@kube1:~# kubeadm init --config=kubeadm-config.yaml --upload-certs
[...]
Your Kubernetes control-plane has initialized successfully!
[...]

Add a comment

Comments (newest first)

No comments yet.