The published chart for 2.9.0 doesn't match the source when enabling Prometheus Operator

[agalue]

I've been deploying Keda with the following configuration to get statistics from it:

prometheus:
  metricServer:
    enabled: true
    podMonitor:
      enabled: true
      namespace: observability
      additionalLabels:
        release: monitor
  operator:
    enabled: true
    podMonitor:
      enabled: true
      namespace: observability
      additionalLabels:
        release: monitor
    prometheusRules:
      enabled: true
      namespace: observability
      additionalLabels:
        release: monitor
      alerts:
      - alert: KedaScalerErrors
        annotations:
          description: Keda scaledObject {{ $labels.scaledObject }} is experiencing errors with {{ $labels.scaler }} scaler
          summary: Keda Scaler {{ $labels.scaler }} Errors
        expr: |
          sum by ( scaledObject , scaler) (rate(keda_metrics_adapter_scaler_errors[2m])) > 0
        for: 2m
        labels:
          namespace: observability # Workaround for https://github.com/prometheus-operator/prometheus-operator/issues/3954

Expected Behavior

The above configuration has been working great up to version 2.8.2.

Actual Behavior

When using the above configuration with the published Helm chart for 2.9.0 (from https://kedacore.github.io/charts), the operator doesn't start, complaining about an unexisting CLI attribute, showing the following error:

❯ kubectl logs -n observability keda-operator-8676696456-g2dcr keda-operator
unknown flag: --metrics-port
Usage of /keda:
      --disable-compression                   Disable response compression for k8s restAPI in client-go.  (default true)
      --health-probe-bind-address string      The address the probe endpoint binds to. (default ":8081")
      --kube-api-burst int                    Set the burst for throttling requests sent to the apiserver (default 30)
      --kube-api-qps float32                  Set the QPS rate for throttling requests sent to the apiserver (default 20)
      --kubeconfig string                     Paths to a kubeconfig. Only required if out-of-cluster.
      --leader-elect                          Enable leader election for controller manager. Enabling this will ensure there is only one active controller manager.
      --metrics-bind-address string           The address the metric endpoint binds to. (default ":8080")
      --metrics-service-bind-address string   The address the gRPRC Metrics Service endpoint binds to. (default ":9666")
      --zap-devel                             Development Mode defaults(encoder=consoleEncoder,logLevel=Debug,stackTraceLevel=Warn). Production Mode defaults(encoder=jsonEncoder,logLevel=Info,stackTraceLevel=Error)
      --zap-encoder encoder                   Zap log encoding (one of 'json' or 'console') (default )
      --zap-log-level level                   Zap Level to configure the verbosity of logging. Can be one of 'debug', 'info', 'error', or any integer value > 0 which corresponds to custom debug levels of increasing verbosity (default )
      --zap-stacktrace-level level            Zap Level at and above which stacktraces are captured (one of 'info', 'error', 'panic').
unknown flag: --metrics-port
      --zap-time-encoding time-encoding       Zap time encoding (one of 'epoch', 'millis', 'nano', 'iso8601', 'rfc3339' or 'rfc3339nano'). Defaults to 'epoch'. (default )

Interestingly, the source code has the correct setting:

https://github.com/kedacore/charts/blob/v2.9.0/keda/templates/12-keda-deployment.yaml#L76

However, what you get after installing the chart seems to show something from older versions, leading to the problem.

Steps to Reproduce the Problem

1. Add the Helm repository: helm repo add kedacore https://kedacore.github.io/charts
2. Update local repositories: helm repo update
3. Download the chart code from the repo helm pull kedacore/keda
4. Uncompress the downloaded file tar xvzf keda-2.9.0.tgz
5. Open the deployment file (look around line 76): vim keda/templates/12-keda-deployment.yaml
6. The downloaded template references --metrics-port instead of --metrics-bind-address

Specifications

• KEDA Version: 2.9.0
• Platform & Version: Minikube
• Kubernetes Version: 1.25.3
• Scaler(s): N/A

[rossedman]

This happened to me as well, can confirm the steps above

[zroubalik]

Thanks for reporting, we will release a new version next week to fix this.

[zroubalik]

This is a duplicate of this: #342