keda: STAN scaler does not scale from 0

A clear and concise description of what the bug is.

Expected Behavior

When there messages in the queue and the HPA is scaled down to 0, the scaler should create at least one instance of a pod to create the subscription with STAN.

Actual Behavior

When the loop runs, The logic checks if the pending count of messages is greater than 0 and that the queue names match. It never takes into account if the subscription does not exist.

Steps to Reproduce the Problem

1. Create a new ScaledObject that can scale to 0

apiVersion: keda.k8s.io/v1alpha1
kind: ScaledObject
metadata:
  name: test-scaledobject
  labels:
    deploymentName: test
spec:
  pollingInterval: 10   # Optional. Default: 30 seconds
  cooldownPeriod: 30   # Optional. Default: 300 seconds
  minReplicaCount: 0   # Optional. Default: 0
  maxReplicaCount: 5  # Optional. Default: 100  
  scaleTargetRef:
    deploymentName: test
  triggers:
  - type: stan
    metadata:
      natsServerMonitoringEndpoint: "stan.default:8222"
      queueGroup: "place_indexer"
      durableName: "ImDurable"
      subject: myQueue
      lagThreshold: "100" 

1. cURL the STAN channel endpoint. It will return a response like:

{
  "name": "myQueue",
  "msgs": 500,
  "bytes": 0,
  "first_seq": 0,
  "last_seq": 0,
  "subscriptions": []

1. Keda will ignore the msgs field because there are no active subscriptions.
2. If minReplicaCount is set to 1, scaling works normally.

Specifications

• Version: 1.0.0
• Platform: Linux
• Scaler(s): STAN