Could you please help me to understand in what cases I need to configure --entrypoints.web.address=:80 additional argument if I am running Traefik in kubernetes? It works totally fine without this additional argument but in most examples all over the Internet people are using this configuration.
Would you please let us know how did you deploy Traefik on K8S Cluster? Did you use the official helm Chart or did you deploy it manually?
Entrypoints are required parameters to open a specific port and allow Traefik to listen on that specific port. That port is an entry point for incoming network traffic to the Traefik instance, it might be TCP or UDP.
If you haven't specified the entrypoint manually, probably it has been done automatically depending of the way how you deploy Traefik on a cluster. If you use the Helm chart it will add entrypoints as CLI arguments.
You can describe the pod with Traefik to see what are available entrypoints. Then those ports are mapped as a targetPort in a K8S service to expose Traefik instance outside Kubernetes cluster.
Thank you for your reply. Yes, I am using Helm to deploy Traefik to k8s. What I've figured out that it is better not to define them, as otherwise I need to update ports section correspondingly in values.yaml file.
It all depends on how you deploy Traefik. If you use the Helm Chart most of the work is already coded into templates, so you don't have to configure it explicitly in config files. I encourage you to review the official Helm Chart repo to see how it is configured.
Comparing to the manual deployment, you have to configure everything by yourself including CRD's, healtcheck probes, CLI args, pvc and so on...
In both cases, it is good to have the source of truth for your configuration in Git repo and deploy the configuration to your cluster using any automation solutions (gitops approach).