Fix e2e tests - pod already exists intermittent issue

[sandeepkunusoth]

This PR fixes intermittent E2E test failures caused by kubectl run attempting to create a pod with same name (client) that may already exist from previous test runs.
failed job link: https://github.com/valkey-io/valkey-operator/actions/runs/25645368225/job/75273172729
Error:
[FAILED] Timed out after 120.001s. The function passed to Eventually failed at /home/runner/work/valkey-operator/valkey-operator/test/e2e/metrics_exporter_test.go:386 with: Unexpected error: <*fmt.wrapError | 0x29b7c393800>: "kubectl run client --image=valkey/valkey:9.0.0 --restart=Never -- valkey-cli -c -h valkeycluster-no-exporter.default.svc.cluster.local ACL LIST" failed with error "Error from server (AlreadyExists): pods \"client\" already exists\n": exit status 1 { msg: "\"kubectl run client --image=valkey/valkey:9.0.0 --restart=Never -- valkey-cli -c -h valkeycluster-no-exporter.default.svc.cluster.local ACL LIST\" failed with error \"Error from server (AlreadyExists): pods \\\"client\\\" already exists\\n\": exit status 1",

Changes

• Updated E2E tests to use unique client pod names or ensure cleanup before pod creation.
• Prevents failures like:
  Error from server (AlreadyExists): pods "client" already exists

Testing

• Verified tests no longer fail intermittently due to existing client pods.

Checklist

• This Pull Request is related to one issue.
• Commit message explains what changed and why
• Tests are added or updated.
• I have run pre-commit locally (pre-commit run --all-files or hooks on commit)