Description of problem:

When configuring a server for the model serving feature, The user can decide to not tick "Require token authentication" (it's actually the default choice); In this case, after deploying a model, it won't be possible to receive any inference from the provided endpoint. Requests will receive the Openshift "Application is not available" error page as a response.

Prerequisites (if any, like setup, operators/versions):

latest model serving live build

Steps to Reproduce

1. Create DSP
2. Configure Server (without token authentication)
3. Deploy model
4. Try to reach the inference endpoint

Actual results:

The Openshift "Application is not available" error page is received as a response from the inference endpoint

Expected results:

Any user is able to reach the inference endpoint and get an inference result back, without using any token

Reproducibility (Always/Intermittent/Only Once):

always

Build Details:

Workaround:

Enable token authentication, include token in request to inference endpoint

Additional info: