The storage suite reuses pod manifests with static names between test cases. Because the test cases are executed in the same namespace, the pod from one case may not have been garbage collected prior to the next test case creating the pod.  Pod creation then fails due to name collision.

stderr:
Error from server (AlreadyExists): error when creating "./assets/reboot/pod-with-pvc.yaml": pods "test-pod" already exists

Hard to say as it depends on the garbage collection timing.

1. cd ~/microshift/test
2. ./bin/scenario.sh create el94-src@storage.sh
3. ./bin/scenario.sh run el94-src@storage.sh 
4. If test fails, examine the RF log output.  Cases are not run in a predetermined order, which might affect the likelihood of the flake.

Test fails during test case setup due to pod name collision

The pod should never fail to create because of a name collision

For a recent example, see: