Enable Advanced Metrics in Retina with Existing Hubble/Cilium Setup

[angelbarrera92]

Is your feature request related to a problem? Please describe.
I'm trying to enable advanced metrics ("advance" mode) in Retina on top of an already existing Hubble/Cilium setup. I’m unable to get the detailed metrics exposed, and it's unclear how to properly integrate Retina into a pre-existing Cilium + Hubble deployment. I’m particularly frustrated because although the Helm chart appears to support disabling Hubble, doing so seems to prevent Retina from functioning as expected.

Describe the solution you'd like
Clear documentation or guidance on how to deploy Retina in “advance” mode in an environment where Cilium and Hubble are already installed and configured independently. Specifically, I want to understand:

• How to configure Retina to leverage the existing Hubble metrics pipeline
• Whether I need Hubble components within the Retina chart enabled, or if I can rely solely on the existing ones
• What configuration is needed to expose the detailed metrics Retina is capable of

Describe alternatives you've considered
I tried:

1. Deploying the retina-hubble Helm chart with all internal Hubble components disabled:

hubble:
  enabled: false
  ui:
    enabled: false
  tls:
    enabled: false
    auto:
      enabled: false
  relay:
    enabled: false
    tls:
      server:
        enabled: false

2. Pointing the existing Hubble deployment to expose these metrics:

hubble:
  metrics:
    enabled:
      - 'flow:sourceEgressContext=pod;destinationIngressContext=pod'
      - 'tcp:sourceEgressContext=pod;destinationIngressContext=pod'
      - 'dns:query;sourceEgressContext=pod;destinationIngressContext=pod'
      - 'drop:sourceEgressContext=pod;destinationIngressContext=pod'

However, even after these attempts, when I port-forward a retina-agent pod, I still only see the default metrics (like DNS stats, TCP connection stats), and none of the advanced flow-level metrics seem to be present.

Additional context
Here is the full values.yaml used with the chart:

os:
  windows: false
  linux: true
operator:
  enabled: true
  repository: ghcr.io/microsoft/retina/retina-operator
  tag: v0.0.33
agent:
  enabled: true
  repository: ghcr.io/microsoft/retina/retina-agent
  tag: v0.0.33
  init:
    enabled: true
    repository: ghcr.io/microsoft/retina/retina-init
    tag: v0.0.33

enablePodLevel: true
remoteContext: true

logLevel: info
hubble:
  enabled: false
  ui:
    enabled: false
  tls:
    enabled: false
    auto:
      enabled: false
  relay:
    enabled: false
    tls:
      server:
        enabled: false

Command used:

$ helm upgrade --install retina oci://ghcr.io/microsoft/retina/charts/retina-hubble \
  --version v0.0.33 \
  --namespace kube-system \
  -f values-hubble.yaml

Questions:

• Does Retina require Hubble components to be deployed via its Helm chart?
• Can Retina work with an existing Cilium/Hubble setup?
• How can I confirm it's working?