Helm chart: container resources hardcoded at 128Mi, not configurable

[lukacsi]

Problem

The Helm chart hardcodes container resource limits at 128Mi memory in deployment.yaml. There's no resources key in values.yaml to override them.

On a small homelab cluster (3 nodes, ~15 apps, ~60 secrets), the pod consistently OOMs at 128Mi. Controllers with unfiltered Secret sources load all cluster secrets into memory for join evaluation — including TLS certs, CNPG credentials, and service account tokens. An enterprise cluster with hundreds of secrets would be far worse.

128Mi is not a viable default for any real deployment.

Fix

Already implemented in lukacsi/dcontroller@e9f6e01:

# deployment.yaml — was hardcoded, now templatable
{{- with .Values.resources }}
resources:
  {{- toYaml . | nindent 10 }}
{{- end }}

# values.yaml — new default
resources:
  limits:
    cpu: 500m
    memory: 512Mi
  requests:
    cpu: 10m
    memory: 128Mi

Happy to open a PR.