NTO's e2e tests run in a sequence.  It looks like there is a race condition where the previous test has not finished rollback and application of the default "openshift-node" profile and  the new tests started running.  This can create test failures even though the test actually did what it was supposed to do.  An example of this issue is here:

https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/origin-ci-test/logs/periodic-ci-openshift-psap-ci-artifacts-release-4.8-nto-e2e-master/1493722015960403968/artifacts/e2e-master/nightly/artifacts/001__nto__run_e2e_test/nto_e2e_test.log

Address this issue either by either waiting for "openshift-node" profiles being Applied after every test completes or prior to starting a new test.  The former seems a better option at this point.