Description of problem:

The tests for rbac kubernetes plugin are failing on K8s clusters (both AKS and GKE, Helm chart and Operator). The warning in the UI says this:

Error communicating with Kubernetes: FETCH_ERROR, message: request to https://mapt-6d3bjsz1.hcp.westus3.azmk8s.io/api/v1/configmaps?labelSelector=backstage.io%2Fkubernetes-id%3Ddeveloper-hub failed, reason: unable to verify the first certificate

That sounds like it is ignoring the skipTLSVerify: true and I found a bug report at backstage regarding the same: https://github.com/backstage/backstage/issues/28486
And a similar error with Operator:

Error communicating with Kubernetes: FETCH_ERROR, message: request to https://api.rhdh-4-17-us-east-2-qk29k.rhdh-qe.devcluster.openshift.com:6443/api/v1/pods?labelSelector=backstage.io%2Fkubernetes-id%3Ddeveloper-hub failed, reason: self-signed certificate in certificate chain

https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/test-platform-results/logs/periodic-ci-red[…]ta/6d43a70f5aa1dafa8cdfab5ce539002b71312789.zip

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

Steps to Reproduce

1. Run e2e tests (nightly) on anything except the OCP Helm chart.

It fails on:

• • OCP Operator
  • AKS and GKE Helm and Operator

Actual results:

Tests fail with the above error

Expected results:

Tests pass

Reproducibility (Always/Intermittent/Only Once):

Always, since the tests were merged

Build Details:

Additional info (Such as Logs, Screenshots, etc):