Description of problem: The up test for the prometheus pod (https://github.com/openshift/cluster-monitoring-operator/blob/650c320463c15b3f5a64abcd02a066c7a9c15b4a/test/e2e/main_test.go#L87) counts the number of series returned. Due to prometheus staleness notion this will return samples for all series that have received samples in the last 5 minutes. Due to improved availability during 4.10 development a 5 minute interval can yield 4 results instead of the expected two. Since the test only polls this condition for one minute it can time out waiting to see 2 pods. How reproducible: Run the test linked above. Actual results: Test fails with message "wait for prometheus-k8s: expected 2 Prometheus instances but got: 4: timed out waiting for the condition" Expected results: Test passes Additional info: To fix this, use last_over_time aggregation to reduce the lookback period.
LGTM, Test passes
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Important: OpenShift Container Platform 4.11.0 bug fix and security update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2022:5069