Could anyone help me understand why K8S allows users to define multiple port names for a specific port? Is there any valid reason behind such an acceptance pattern?
Originally it was a side-effect of the design, but it turns out some people use it to get different hostPorts (rare) so we decided not to “fix” it.
@thockin : Thanks for the quick response! But do we know the list of use cases that people use today and the recommendation/ good practice from k8s? I have no clue who added this support in k8s and why. It would be good if someone can point me to the bug/PR in the gitHub.
It wasn’t added on purpose, it was the side-effect of other decisions.
Best practice would be “don’t use it if you can’t explain why you need it”. Really, I doubt we would have added it on purpose, but it was emergent and we decided not to “fix” it (breaking users).
I tried but I could not find the issue where this was discussed. I will try again later.