[SPARK-57856] Authenticate minikube GitHub API requests to avoid CI rate limit failures

[dongjoon-hyun]

What changes were proposed in this pull request?

This PR adds GITHUB_TOKEN to the medyagh/setup-minikube steps in
.github/workflows/build_and_test.yml (the k8s-integration-tests and
helm-tests jobs) so that minikube authenticates its GitHub API requests.

Why are the changes needed?

When starting minikube with a pinned kubernetes-version (e.g. 1.36.0),
minikube resolves the exact version by querying the GitHub API
(api.github.com/repos/kubernetes/kubernetes/releases/tags/v1.36.0).
Without a token this request is unauthenticated and limited to 60 requests
per hour per runner IP. Because these jobs fan out across a large matrix
(max-parallel: 20), the shared limit is exhausted and the jobs fail with:

• https://github.com/apache/spark-kubernetes-operator/actions/runs/28527974981/job/84569529838

Exiting due to K8S_FAIL_CONNECT: error fetching Kubernetes version list from GitHub:
GET https://api.github.com/repos/kubernetes/kubernetes/releases/tags/v1.36.0:
403 API rate limit exceeded for 13.71.231.42.

(But here's the good news: Authenticated requests get a higher rate limit.
Check out the documentation for more details.) [rate reset in 41m27s]

Providing GITHUB_TOKEN switches minikube to authenticated requests, raising
the limit to 1,000 requests/hour per repository and removing the flaky failure.

Does this PR introduce any user-facing change?

No. This is a CI-only change.

How was this patch tested?

Existing CI. The k8s-integration-tests and helm-tests jobs run minikube
setup and pass without hitting the GitHub API rate limit.

Was this patch authored or co-authored using generative AI tooling?

Generated-by: Claude Opus 4.8

[dongjoon-hyun]

Thank you, @peter-toth .

[dongjoon-hyun]

Merged to main