-
Bug
-
Resolution: Not a Bug
-
Major
-
None
-
4.14.0
-
None
-
Important
-
No
-
Proposed
-
False
-
If you want to see it in CI: https://prow.ci.openshift.org/view/gs/origin-ci-test/pr-logs/pull/openshift_release/42389/rehearse-42389-periodic-ci-shiftstack-shiftstack-ci-main-periodic-4.14-e2e-openstack-nfv-intel/1691878302945382400
Check here, the masters being updated: https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/origin-ci-test/pr-logs/pull/openshift_release/42389/rehearse-42389-periodic-ci-shiftstack-shiftstack-ci-main-periodic-4.14-e2e-openstack-nfv-intel/1691878302945382400/artifacts/e2e-openstack-nfv-intel/gather-extra/artifacts/oc_cmds/machineconfigpools
Description of problem:
master are rebooted once a PerformanceProfile is being created for CNF workers
Version-Release number of selected component (if applicable):
4.14.0
Steps to Reproduce:
* Install a compact cluster, 3 masters - 0 worker
* Run this command: oc patch Scheduler cluster --type=merge --patch '{ "spec": { "mastersSchedulable": false } }'
* The masters should not have the worker role anymore
* Create a Performance Profile, with node selector set for workers only, not masters
Actual results:
The masters are rebooted.
Expected results:
The masters are not rebooted; only workers would be rebooted if we had any worker, since the PerformanceProfile would be applied.
Additional info:
* Started to happen around 1st of August in our CI * Example of CI job that failed: https://prow.ci.openshift.org/view/gs/origin-ci-test/pr-logs/pull/openshift_release/42389/rehearse-42389-periodic-ci-shiftstack-shiftstack-ci-main-periodic-4.14-e2e-openstack-nfv-intel/1691878302945382400 * https://redhat-internal.slack.com/archives/CH98TDJUD/p1692034558077379
- is caused by
-
-
- Closed
-
- relates to
-
-
- Closed
-
