fix(ci): increase wait kubeconfig and user-authz ready

[universal-itengineer]

Description

Increase the waiting window for nested cluster kubeconfig validation and make the validation diagnostics non-fatal inside the retry loop.

The generated kubeconfig check now captures the kubectl get nodes error output, prints additional authorization diagnostics, and retries for longer while user-authz materializes the ClusterAuthorizationRule into RBAC bindings. The Deckhouse readiness helper also waits for the user-authz module in addition to user-authn.

Why do we need it, and what problem does it solve?

Nested e2e bootstrap can create the service account token and ClusterAuthorizationRule successfully, but fail immediately afterwards when validating the generated kubeconfig. In the observed failure, the ClusterRoleBinding managed by user-authz appeared several minutes after the ClusterAuthorizationRule was applied, while the previous retry window was only about 50 seconds.

This change removes that race from the CI path and keeps useful diagnostics in the logs without failing the job on the first temporary authorization error.

What is the expected result?

1. Run the nested e2e bootstrap pipeline.
2. Wait until Deckhouse, user-authn, and user-authz are ready in the nested cluster.
3. Generate the nested cluster kubeconfig.
4. Verify that kubeconfig validation retries until the generated service account can list nodes, or fails only after the extended timeout with diagnostic output.

Checklist

• The code is covered by unit tests.
• e2e tests passed.
• Documentation updated according to the changes.
• Changes were tested in the Kubernetes cluster manually.

Changelog entries

section: ci
type: fix
summary: Wait longer for nested e2e kubeconfig authorization to become ready.
impact_level: low