Proposal
KEDA should be offer a way of filtering out specific labels when it copies them from a ScaledObject to the HPA.
Use-Case
When you create an ScaledObject using tanka it gets labeled with an environment hash (tanka.dev/environment=<HASH>) to get track of the objects and being able to prune removed objects from your code and so.
As KEDA copies all the labels applied to the ScaledObject to the HPA the tanka's environment label gets propagated so. This causes tanka to continuously show in the diff that the HPA needs to be removed, and, in fact, our deployment pipeline removes it (I have to thank KEDA being that quick on re-creating it, but is not the ideal scenario).
Is this a feature you are interested in implementing yourself?
Maybe
Anything else?
No response
Proposal
KEDA should be offer a way of filtering out specific labels when it copies them from a
ScaledObjectto theHPA.Use-Case
When you create an
ScaledObjectusing tanka it gets labeled with an environment hash (tanka.dev/environment=<HASH>) to get track of the objects and being able to prune removed objects from your code and so.As KEDA copies all the labels applied to the
ScaledObjectto theHPAthe tanka's environment label gets propagated so. This causes tanka to continuously show in the diff that theHPAneeds to be removed, and, in fact, our deployment pipeline removes it (I have to thank KEDA being that quick on re-creating it, but is not the ideal scenario).Is this a feature you are interested in implementing yourself?
Maybe
Anything else?
No response