protocolVersion used by the predictor

[Csehpi]

Hi community,

I am trying to understand how protocolVersion is used by the predictor and I would like to ask some help from you.

Here is a little example:

apiVersion: serving.kserve.io/v1beta1
kind: InferenceService
metadata:
  name: sklearn-from-uri
spec:
  predictor:
    model:
      modelFormat:
        name: sklearn
      protocolVersion: v1
      runtime: kserve-sklearnserver
      storageUri: "gs://kfserving-examples/models/sklearn/1.0/model"

The doc says: Protocol version to use by the predictor (i.e. v1 or v2 or grpc-v1 or grpc-v2)
https://kserve.github.io/website/latest/reference/api/#serving.kserve.io/v1beta1.PredictorExtensionSpec

When pod, service, etc. is ready then I am able to call both V1 and V2 endpoints. Is this the expected behaviour?

Thanks,
Peter