Test flakes with error No API token found for service account "default", retry after the token is automatically created

[ramr]

See: https://storage.googleapis.com/origin-ci-test/pr-logs/pull/18587/test_pull_request_origin_extended_conformance_gce/16274/build-log.txt

The above mentioned PR is hitting it consistently. There are 4 of the same type of failures in the attached log and all the failures are within a couple of seconds of each other.

It looks like the test creates a service account and then tries to create a pod before the API token for the service account is available. It looks like its waiting 10 seconds on that.

Relevant snippet from the logs:

Feb 14 21:17:26.058: INFO: >>> kubeConfig: /tmp/cluster-admin.kubeconfig
STEP: Building a namespace api object
Feb 14 21:17:30.411: INFO: About to run a Kube e2e test, ensuring namespace is privileged
Feb 14 21:17:30.595: INFO: No PodSecurityPolicies found; assuming PodSecurityPolicy is disabled.
STEP: Waiting for a default service account to be provisioned in namespace
[BeforeEach] [k8s.io] Sysctls
  /tmp/openshift/build-rpms/rpm/BUILD/origin-3.9.0/_output/local/go/src/github.com/openshift/origin/vendor/k8s.io/kubernetes/test/e2e/common/sysctl.go:56
[It] should support unsafe sysctls which are actually whitelisted [Suite:openshift/conformance/parallel] [Suite:k8s]
  /tmp/openshift/build-rpms/rpm/BUILD/origin-3.9.0/_output/local/go/src/github.com/openshift/origin/vendor/k8s.io/kubernetes/test/e2e/common/sysctl.go:101
STEP: Creating a pod with the kernel.shm_rmid_forced sysctl
Feb 14 21:17:40.101: INFO: Unexpected error occurred: No API token found for service account "default", retry after the token is automatically created and added to the service account

[smarterclayton]

Probably introduced by #18508

/assign @deads2k

[smarterclayton]

Created a bug since this may break end user apps https://bugzilla.redhat.com/show_bug.cgi?id=1546466

[deads2k]

fixed

[liggitt]

fixed in #18661