[BUG] copy-azurekeyvault-env container gets default CPU requests of 50m

[nemobis]

Note: Make sure to check out known issues (https://akv2k8s.io/troubleshooting/known-issues/) before submitting

Components and versions

1.3.1

Init Containers:
  copy-azurekeyvault-env:
    Container ID:  containerd://c0c88b79f5be181f5ca3d99636a9a4cec6f77f2d020328d01d76796eec8910d0
    Image:         kubeplatformproduction.azurecr.io/spvest/azure-keyvault-env:1.3.1
    Image ID:      kubeplatformproduction.azurecr.io/spvest/azure-keyvault-env@sha256:ebc3bf9308b85a3a17924db774801fc47bc77c35e8d50013baeb939c5b440bc3
    Port:          <none>
    Host Port:     <none>
    Command:
      sh
      -c
      cp /usr/local/bin/azure-keyvault-env /azure-keyvault/
    State:          Terminated
      Reason:       Completed
      Exit Code:    0
      Started:      Fri, 12 Jan 2024 22:37:48 +0200
      Finished:     Fri, 12 Jan 2024 22:37:48 +0200
    Ready:          True
    Restart Count:  0
    Limits:
      cpu:     50m
      memory:  100Mi
    Requests:
      cpu:        50m
      memory:     100Mi
    Environment:  <none>
    Mounts:
      /azure-keyvault/ from azure-keyvault-env (rw)
      /var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-jcdsd (ro)

Describe the bug

The copy-azurekeyvault-env container gets 50m in CPU requests (possibly due to relying on defaults #210) despite needing an infinitesimal amount of CPU (I believe less than 0.00003 %, as it's less than anything else in my monitoring).

To Reproduce
Steps to reproduce the behavior: use is, check kubectl describe pod.

Expected behavior
Containers should provide realistic CPU and memory requests, to avoid reserving resources they don't need.

Additional context
A 50m CPU request means that, on a 2-core node in AKS, where 1900m are allocatable but some 500m are used by kube-system (cf. Azure/AKS#3525), it constantly costs almost 4 % of the node, or some 2.5 €/month for one of the cheapest VM types, despite being active only few seconds.