Update docker image used

I have noticed that you still use the image `aledbf/kube-keepalived-vip` from upstream, which hasn't been updated since 2019

How about you automatically build the image in your CI and publish it on ghcr ? I think an image running in all nodes, privileged, without network isolation, and with read access to the cluster should be updated