minReplicaCount not applied on a trigger type: cron

[benoit-parmentier-2]

Report

Hello,

On my ScaledObject, I configured a type: cron trigger.
The cron type trigger allows me to scale to 0, as shown here: https://keda.sh/docs/2.13/scalers/cron/#scale-to-0-during-off-hours.

• I would like outside of the 7-21 range to have the minReplicaCount:0 applied.
• I would like in the range 7-21, other triggers to be used, for example type: cpu or type: datadog

Expected Behavior

1) With the configuration below scaledobject-cron-only.yaml.

It works as expected.

• Between start and end, the desiredReplicas: 2
• Outside of start and end, the minReplicaCount: 0 is applied

spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: nuxtjs-helloworld-green
  minReplicaCount: 0
  triggers:
    - type: cron
      metadata:
        timezone: "Europe/Paris"
        start: "0 7 * * *"
        end: "30 10 * * *"
        desiredReplicas: "2"

2) With the configuration below scaledobject-cron-memory.yaml

It works as expected.

• Between start and end, the desiredReplicas: 2, plus autoscaling works well on memory
• Outside of start and end, the minReplicaCount: 0 is applied

  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: nuxtjs-helloworld-green
  minReplicaCount: 0
  pollingInterval: 30
  cooldownPeriod: 3
  maxReplicaCount: 5
  fallback:
    failureThreshold: 3
    replicas: 5
  triggers:
    - type: memory
      metricType: Utilization
      metadata:
        value: "1"
    - type: cron
      metadata:
        timezone: "Europe/Paris"
        start: "0 7 * * *"
        end: "30 10 * * *"
        desiredReplicas: "2"

3) With the configuration below scaledobject-cron-datadog.yaml

It doesn't work as expected.

• Between start and end, the desiredReplicas: 2
• Outside of start and end, the minReplicaCount: 0 is not applied. There is a pod that is created.

  triggers:
    - type: cron
      metadata:
        timezone: "Europe/Paris"
        start: "0 7 * * *"
        end: "30 10 * * *"
        desiredReplicas: "2"
    - authenticationRef:
        kind: ClusterTriggerAuthentication
        name: keda-datadog
      metadata:
        age: "120"
        lastAvailablePointOffset: "1"
        metricUnavailableValue: "1"
        query: sum:nginx.net.request_per_s{env:preprod,service:nuxtjs-helloworld}
        queryValue: "100"
        timeWindowOffset: "30"
      metricType: AverageValue
      type: datadog

Could you explain to me why configuration 3 does not enforce the minReplicaCount: 0 ?

Thanks a lot,
Benoît

Actual Behavior

As stated in my 3) point, when I put a type:datadog trigger with a trigger type:cron, it doesn't work.

Steps to Reproduce the Problem

As stated in my 3) point, you can use the configuration below.

  triggers:
    - type: cron
      metadata:
        timezone: "Europe/Paris"
        start: "0 7 * * *"
        end: "30 10 * * *"
        desiredReplicas: "2"
    - authenticationRef:
        kind: ClusterTriggerAuthentication
        name: keda-datadog
      metadata:
        age: "120"
        lastAvailablePointOffset: "1"
        metricUnavailableValue: "1"
        query: sum:nginx.net.request_per_s{env:preprod,service:nuxtjs-helloworld}
        queryValue: "100"
        timeWindowOffset: "30"
      metricType: AverageValue
      type: datadog

Logs from KEDA operator

example

KEDA Version

2.13.0

Kubernetes Version

1.28

Platform

Microsoft Azure

Scaler Details

No response

Anything else?

We use Keda version 2.13.2 (not possible to select this version in KEDA Version dropdown list)