🐛 Agents stop update managedcluster status when clock is out of sync.

[xuezhaojun]

Summary

This patch aim to stop agents from updating managed cluster status when clock is already out of sync.

The issue occurs when the hub sets the managed cluster status to unknown due to an outdated lease, while the agent simultaneously sets the managed cluster status to available because the API server is accessible. This conflict causes the managed cluster status to flap back and forth frequently, triggering numerous reconciliations and resulting in performance issues.

[codecov]

Codecov Report

Attention: Patch coverage is 54.54545% with 5 lines in your changes missing coverage. Please review.

Project coverage is 63.77%. Comparing base (17e05da) to head (528c61e).
Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
...stration/spoke/managedcluster/status_controller.go	33.33%	3 Missing and 1 partial ⚠️
pkg/registration/hub/lease/clocksynccontroller.go	80.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #770      +/-   ##
==========================================
+ Coverage   63.60%   63.77%   +0.16%     
==========================================
  Files         192      192              
  Lines       18433    18596     +163     
==========================================
+ Hits        11725    11860     +135     
- Misses       5735     5757      +22     
- Partials      973      979       +6     

Flag	Coverage Δ
unit	63.77% <54.54%> (+0.16%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jgato]

what are the consequences of not been able to update the ManagedCluster status? Do we lost any kind of feature?
It would happen that we would need to have different clocks. Like a managedcluster, testing different functionalities related with time, ptp, ntp, and therefore, it would be "playing" with the clock.

[xuezhaojun]

what are the consequences of not been able to update the ManagedCluster status? Do we lost any kind of feature? It would happen that we would need to have different clocks. Like a managedcluster, testing different functionalities related with time, ptp, ntp, and therefore, it would be "playing" with the clock.

@jgato The OCM is not designed to work under the situation when clocks are not synced, if clocks are not synced, the first priority is to fix it.

This PR prevents status flapping by ensuring managed clusters remain in Unknown state instead of alternating between Unknown and Available.

[xuezhaojun]

/assign @qiujian16

[xuezhaojun]

/assign @elgnay

[qiujian16]

We need also document this change in the website.

[qiujian16]

/approve

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: qiujian16, xuezhaojun

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• pkg/registration/OWNERS [qiujian16,xuezhaojun]
• test/OWNERS [qiujian16]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[xuezhaojun]

We need also document this change in the website.

Agree, will add tasks for it.

[xuezhaojun]

@qiujian16 After run agent again, check if conditions become True, please take another look.

[xuezhaojun]

@qiujian16 A task of updating doc is added: open-cluster-management-io/open-cluster-management-io.github.io#443