Uploaded image for project: 'Red Hat Advanced Cluster Management'
  1. Red Hat Advanced Cluster Management
  2. ACM-22000

[Tech Debt] Enhance the manifestWork deletion process to fix resources leftover issue after cluster is deleted

XMLWordPrintable

    • Icon: Epic Epic
    • Resolution: Done
    • Icon: Major Major
    • MCE 2.10.0
    • MCE 2.9.0
    • Server Foundation
    • None
    • [Tech Debt] Enhance the manifestWork deletion process to fix resources leftover issue after cluster is deleted
    • Quality / Stability / Reliability
    • 3
    • False
    • Hide

      None

      Show
      None
    • False
    • Not Selected
    • Done
    • 0% To Do, 0% In Progress, 100% Done
    • Important

      During the managedcluster-import-controller e2e test, we met the  random issue that the klusterlet CRD and CR are not deleted after the cluster and the manifestWorks are all deleted. 

      After debug ,we found the import-controller force deletes the klusterlet-CRD manifestWork, but work-agent did not receive the klusterlet-CRD manifestWork update event to trigger deletion in some cases. 

      We decide to add a deleting condition to the manifestWork by the work-agent after watch the manifestWork is deleting.  The import-controller force deletes the klusterlet-CRD manifestWork after the mainfestWork has this deleting condition.

       

              jqiu@redhat.com Jian Qiu
              zyin@redhat.com Zhiwei Yin
              Hui Chen Hui Chen
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated:
                Resolved: