Bug 1731263 - e2e flake: [sig-scheduling] SchedulerPreemption
Summary: e2e flake: [sig-scheduling] SchedulerPreemption
Status: POST
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: kube-scheduler
Version: 4.2.0
Hardware: Unspecified
OS: Unspecified
Target Milestone: ---
: 4.3.0
Assignee: Mike Dame
QA Contact: ge liu
Whiteboard: buildcop
: 1733005 1751661 (view as bug list)
Depends On:
Blocks: 1751626
TreeView+ depends on / blocked
Reported: 2019-07-18 19:49 UTC by Matthew Staebler
Modified: 2020-01-09 08:28 UTC (History)
9 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Last Closed:
Target Upstream Version:

Attachments (Terms of Use)

System ID Priority Status Summary Last Updated
Github openshift origin pull 23050 None None None 2019-08-13 15:16:24 UTC
Github openshift origin pull 23645 None None None 2019-08-21 17:01:08 UTC
Github openshift origin pull 23758 None None None 2019-09-10 15:13:50 UTC

Description Matthew Staebler 2019-07-18 19:49:49 UTC
[sig-scheduling] SchedulerPreemption [Serial] validates lower priority pod preemption by critical pod [Suite:openshift/conformance/serial] [Suite:k8s]

[sig-scheduling] SchedulerPreemption [Serial] validates basic preemption works [Suite:openshift/conformance/serial] [Suite:k8s]

        ErrStatus: {
            TypeMeta: {Kind: "", APIVersion: ""},
            ListMeta: {SelfLink: "", ResourceVersion: "", Continue: ""},
            Status: "Failure",
            Message: "pods \"pod0-sched-preemption-low-priority\" not found",
            Reason: "NotFound",
            Details: {
                Name: "pod0-sched-preemption-low-priority",
                Group: "",
                Kind: "pods",
                UID: "",
                Causes: nil,
                RetryAfterSeconds: 0,
            Code: 404,

See https://bugzilla.redhat.com/show_bug.cgi?id=1700458 for where the test failed previously in 4.1.

Comment 3 Mike Dame 2019-07-31 18:24:09 UTC
*** Bug 1733005 has been marked as a duplicate of this bug. ***

Comment 4 ravig 2019-08-13 14:29:58 UTC
Origin PR: https://github.com/openshift/origin/pull/23050

Comment 5 Maciej Szulik 2019-08-21 12:43:45 UTC
Mike another one, please.

Comment 6 Mike Dame 2019-08-21 17:01:53 UTC
Updated PR: https://github.com/openshift/origin/pull/23645

Comment 8 Xingxing Xia 2019-09-05 10:06:00 UTC
Hi, the quickest way to move browse https://testgrid.k8s.io/redhat-openshift-release-blocking#redhat-release-openshift-origin-installer-e2e-aws-serial-4.2&sort-by-flakiness results of runs. But search "SchedulerPreemption" there, only saw the case "SchedulerPreemption [Serial] validates pod anti-affinity works in preemption", no the cases of comment 0:
validates lower priority pod preemption by critical pod
validates basic preemption works

Why? How to check the fix effect then if not to manually run openshift-tests?

Comment 9 Xingxing Xia 2019-09-10 09:25:29 UTC
The tested cases "validates lower priority pod preemption by critical pod" and "validates basic preemption works" originally reported by this bug are seen in 4.1 job https://testgrid.k8s.io/redhat-openshift-release-blocking#redhat-release-openshift-origin-installer-e2e-aws-serial-4.1&sort-by-flakiness testing results.
But are not seen in 4.2 job https://testgrid.k8s.io/redhat-openshift-release-blocking#redhat-release-openshift-origin-installer-e2e-aws-serial-4.2&sort-by-flakiness .
I also checked other platforms' jobs, didn't see the cases shown either.
Not sure why. This is confusing, due to the fix landed in 4.2. Thus assigning back to confirm if the cases are disabled or skipped.

Comment 10 Mike Dame 2019-09-10 15:26:04 UTC
I've opened https://github.com/openshift/origin/pull/23758 to reenable these tests

Comment 12 Maciej Szulik 2019-11-05 11:13:41 UTC
*** Bug 1751661 has been marked as a duplicate of this bug. ***

Comment 13 Mike Dame 2019-11-07 19:21:03 UTC
From what I can tell, and as I commented here: https://github.com/openshift/origin/pull/23758#issuecomment-551215646, these tests should be re-enabled since they are no longer marked flaky. I can't find any runs for them, but I'll switch it to QA hoping you guys can see it. If not we need to dig deeper to find out why it isn't running

Comment 15 Mike Dame 2019-11-07 20:00:18 UTC
Turns out it was disabled here: https://github.com/openshift/origin/commit/55358f73aa477f401445a47620dea7e81b27bce9 I reopened my PR above to remove those lines and see how the test runs now

Note You need to log in before you can comment on or make changes to this bug.