autoscaling.keda.sh/paused-replicas: "0" not working

[sharique-masood]

Report

The annotation autoscaling.keda.sh/paused-replicas: "0" is not behaving as expected. Even after applying this annotation to the KEDA ScaledObject, deployments are continuing to get scaled in.

Expected Behavior

The autoscaling.keda.sh/paused-replicas annotation with a value of "0" should prevent any scaling actions (scaling in) from occurring. This should scale down the deployment to zero or 1.

Actual Behavior

Deployments are still being scaled in even after applying the annotation.

Steps to Reproduce the Problem

1. Create or use an existing KEDA ScaledObject.
2. Add the annotation autoscaling.keda.sh/paused-replicas: "0" to the ScaledObject definition, intending to prevent scaling down.
3. Observe that deployments continue to get scaled in, despite the annotation indicating a value of "0."

Logs from KEDA operator

1.6952876629502664e+09	DEBUG	controller.scaledobject	ScaledObject is defined correctly and is ready for scaling	{"reconciler group": "keda.sh", "reconciler kind": "ScaledObject", "name": "" ........

KEDA Version

2.6.1

Kubernetes Version

1.24

Platform

Google Cloud

Scaler Details

No response

Anything else?

No response

[ianebot]

Hi @sharique-masood,

I think the pause feature was introduced in KEDA 2.7.0: https://github.com/kedacore/keda/blob/main/CHANGELOG.md#new-7
You are using KEDA 2.6.1 if I am not mistaken, so you may need to upgrade.

Also keep in mind that there was a bug fix in KEDA 2.11.1 regarding this feature. So you may want to check this out: https://github.com/kedacore/keda/blob/main/CHANGELOG.md#fixes-2

Regards.

[zroubalik]

@ianebot is correct, closing as this is not a bug.