• MCO Sprint 26
    • Moderate
    • -
    • No

      Description of problem:

      The MCO is currently using an old version of the CMO dependency. This impacts the logic that updates the cluster-monitoring-config config map with additional configuration related to ACM Observability.  Such config map now has many more configuration that are reset to their type's zero-value when the CMO tries to marshal the configuration to text. For example, the retentionSize field gets zeroed because it doesn't exist in the current version of the MCO that we import.

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

      2.11, 2.10, 2.9

      How reproducible: 

      Steps to Reproduce:

      1. Create an ACM deployment with at least one spoke. Could be local cluster. 
      2. Configure a custom retentionSize in the CMO stack through the cluster-monitoring-configuration config map.
      3. Deployed ACM Observability
      4. Watch the cluster-monitoring-configuration config map lose the custom retentionSize configuration.

      Actual results:

      retentionSize gets deleted.

      Expected results:

      retentionSize, and any other field that's not related to what MCO needs to customize, stay the same.

      Additional info:

            [ACM-12213] Upgrade CMO dependency in the MCO

            confirmed by Dev, the above extra parameters are fine, close this issue.

            ChangLiang Qu added a comment - confirmed by Dev, the above extra parameters are fine, close this issue.

            pgough@redhat.com rh-ee-tmange  I set up one 2.12 environment today to verify the issue https://issues.redhat.com/browse/ACM-12213, I added some parameters and values in the configmap cluster-monitoring-configuration, like
            data:
            config.yaml: |
            prometheusK8s:
            retention: 24h
            retentionSize: 10GB
             
            They still exist and are not modified once the Observability is deployed. I think the issue is fixed.
            But one question, when the Observability is deployed, there are some extra parameters without values are added in this cm, is it expected?
             
            Another question is this issue planned to be shipped by 2.11.1? smeduri1@redhat.com rhn-support-xiyin rhn-support-vboulos rhaigner@redhat.com FYA.
             

            ChangLiang Qu added a comment - pgough@redhat.com rh-ee-tmange   I set up one 2.12 environment today to verify the issue https://issues.redhat.com/browse/ACM-12213 , I added some parameters and values in the configmap cluster-monitoring-configuration, like data: config.yaml: | prometheusK8s: retention: 24h retentionSize: 10GB   They still exist and are not modified once the Observability is deployed. I think the issue is fixed. But one question, when the Observability is deployed, there are some extra parameters without values are added in this cm, is it expected?   Another question is this issue planned to be shipped by 2.11.1? smeduri1@redhat.com rhn-support-xiyin rhn-support-vboulos rhaigner@redhat.com FYA.  

            Jacob Berger added a comment - Pr for 2.9: https://github.com/stolostron/rhacm-docs/pull/6559/files     PR for 2.10: https://github.com/stolostron/rhacm-docs/pull/6558  

            Jacob Berger added a comment - file changed for this issue: https://github.com/stolostron/rhacm-docs/blob/2.11_stage/release_notes/known_issues_observability.adoc  

            Jacob Berger added a comment - PR for 2.10: https://github.com/stolostron/rhacm-docs/pull/6558  

            Jacob Berger added a comment - See PR: https://github.com/stolostron/rhacm-docs/pull/6542  

              pgough@redhat.com Philip Gough
              rh-ee-doolivei Douglas Camata (Inactive)
              Xiang Yin Xiang Yin
              Votes:
              0 Vote for this issue
              Watchers:
              6 Start watching this issue

                Created:
                Updated:
                Resolved: