Bug 2018222 - cluster-monitoring-operator produces more watch requests than expected
Summary: cluster-monitoring-operator produces more watch requests than expected
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: Monitoring
Version: 4.10
Hardware: Unspecified
OS: Unspecified
unspecified
medium
Target Milestone: ---
: 4.10.0
Assignee: Sunil Thaha
QA Contact: Junqi Zhao
URL:
Whiteboard:
Depends On:
Blocks: 2026802
TreeView+ depends on / blocked
 
Reported: 2021-10-28 14:27 UTC by Adam Kaplan
Modified: 2022-01-11 06:12 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2021-11-12 16:42:48 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)

Description Adam Kaplan 2021-10-28 14:27:07 UTC
Description of problem:

CI tests for 4.10 are flaking with the cluster-monitoring operator reporting too many watches:

[sig-arch][Late] operators should not create watch channels very often [Suite:openshift/conformance/parallel] 

"Operator \"cluster-monitoring-operator\" produces more watch requests than expected: watchrequestcount=69, upperbound=66, ratio=1.0454545454545454"

Version-Release number of selected component (if applicable): 4.10


How reproducible: Sometimes


Steps to Reproduce:

Mainly non-IPI tests seem to be impacted (example - 4.10 aws-upgrade)

Actual results:

[sig-arch][Late] operators should not create watch channels very often [Suite:openshift/conformance/parallel] test fails


Expected results:

Tests pass


Additional info:

https://prow.ci.openshift.org/view/gs/origin-ci-test/pr-logs/pull/openshift_release/22451/rehearse-22451-pull-ci-openshift-builder-release-4.10-openshift-e2e-aws-builds-techpreview/1452970495040294912

Comment 1 Jan Fajerski 2021-10-29 06:35:28 UTC
The fix for https://bugzilla.redhat.com/show_bug.cgi?id=2016352 might improve this.

Comment 2 Simon Pasquier 2021-11-12 16:42:48 UTC
IIUC this is a generic issue with this e2e test checking that operators don't create too many watches and nothing specific to the cluster monitoring operator.

The following PRs should improve the situation:
https://github.com/openshift/origin/pull/26583
https://github.com/openshift/origin/pull/26601


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