Provide support for preventing interference of autoscaling a single scale target when creating a new ScaledJob/ScaledObject

[tomkerkhove]

Proposal

We always recommend people to follow this guidance:

• Only have one ScaledJob/ScaledObject with multiple triggers, instead of multiple instances
• Do not mix and match ScaledJob/ScaledObject with your own HPA

However, we don't provide any prevention for this.

KEDA should automatically be rejecting new ScaledObject resources that are created which target a scale target which is already being autoscaled.

This should cover the following scenarios:

• ScaledJob/ScaledObject targeting workload A, but workload A is already scaled by another ScaledJob/ScaledObject
• ScaledJob/ScaledObject targeting workload A, but workload A is already scaled by another Horizontal Pod Autoscaler

Use-Case

Prevent end-users from having multiple actors scale workloads, either by accident or on purpose, to prevent flapping due to different criteria.

Anything else?

See #3023 for the discussion

[tomkerkhove]

Just to be sure; are we effectively checking this today or was it closed by accident? (I think it's fine but just checking)

[JorTurFer]

It was closed automatically because the feature is fully implemented.
Feature + Docs + Helm chart changes are merged (helm changes merge closed this)