Liqo-fabric port 8081 already in use with Kubespray cluster

[andreadetti]

What happened:

liqo-fabric pods crash on node different from master node

What you expected to happen:

liqo-fabric run

How to reproduce it (as minimally and precisely as possible):

Install a cluster with kubespray and Calico

Anything else we need to know?:

The problem is that kubespray use nginx-proxy on port 8081 and host-network as reverse proxy for accessing API server from nodes different from the master. liqo-fabric pod uses the same port as from the next log:

k logs liqo-fabric-4cnkx -n liqo
I0108 11:14:09.065026       1 unix.go:37] Kernel version 6.8.0-50-generic satisfies the minimum requirements 5.10.0
Error: unable to create manager: error listening on :8081: listen tcp :8081: bind: address already in use

Environment:

• Liqo version: v1.0.0-rc.3
• Liqoctl version: v1.0.0-rc.3
• Kubernetes version (use kubectl version): v1.31.3
• Cloud provider or hardware configuration: bare-metal
• Node image:
• Network plugin and version: Calico
• Install tools: Kubespray
• Others:

[fra98]

Hi @andreadetti, the 8081 port is used for the health probe, but can be changed by setting the --health-probe-bind-address flag.
You can either set it directly or at install-time setting the proper Helm value, i.e. --set networing.fabric.pod.extraArgs={--health-probe-bind-address=:PORT}

[cheina97]

@andreadetti if you are not planning to use the health probe, I suggest to set the flag to "0" in order to disable it