[Bug] GrafanaAlertRuleGroup changes when applied

[AlexEndris]

Describe the bug
Deploying a GrafanaAlertRuleGroup in a k8s cluster using ArgoCD, due to applying defaults (see diff in screenshot) the resource changes and is flagged as "OutOfSync" by ArgoCD.

Version
v5.8.1

To Reproduce
Apply a GrafanaAlertRuleGroup that has been exported by Grafana:

apiVersion: grafana.integreatly.org/v1beta1
kind: GrafanaAlertRuleGroup
metadata:
  name: grafanaalertrulegroup-sample
spec:
  folderRef: test-folder
  instanceSelector:
    matchLabels:
      dashboards: "grafana"
  interval: 5m
  rules:
    - uid: cdi3wts6ubqbkc
      title: Test
      condition: C
      data:
        - refId: A
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: Prometheus
          model:
            disableTextWrap: false
            editorMode: builder
            expr: alertmanager_alerts{instance="10.244.1.21:9093"}
            fullMetaSearch: false
            includeNullMetadata: true
            instant: true
            intervalMs: 1000
            legendFormat: __auto
            maxDataPoints: 43200
            range: false
            refId: A
            useBackend: false
        - refId: B
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: __expr__
          model:
            conditions:
                - evaluator:
                    params: []
                    type: gt
                  operator:
                    type: and
                  query:
                    params:
                        - B
                  reducer:
                    params: []
                    type: last
                  type: query
            datasource:
                type: __expr__
                uid: __expr__
            expression: A
            intervalMs: 1000
            maxDataPoints: 43200
            reducer: last
            refId: B
            type: reduce
        - refId: C
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: __expr__
          model:
            conditions:
                - evaluator:
                    params:
                        - 0
                    type: gt
                  operator:
                    type: and
                  query:
                    params:
                        - C
                  reducer:
                    params: []
                    type: last
                  type: query
            datasource:
                type: __expr__
                uid: __expr__
            expression: B
            intervalMs: 1000
            maxDataPoints: 43200
            refId: C
            type: threshold
      noDataState: NoData
      execErrState: Error
      for: 5m
      annotations:
        description: ""
        runbook_url: ""
        summary: ""
      labels:
        "": ""
      isPaused: false

Expected behavior
Defaults like to: 0 from the "relativeTimeRange" or isPaused: false or interval: 5m don't change.

Actual behavior
to: 0 and isPaused: false vanish, probably due to being defaults, and interval: 5m changes to interval: 5m0s.

Suspect component/Location where the bug might be occurring
Please provide this if you know where this bug might occur otherwise leave as unknown

Screenshots
Left is the changed resource, right is the actual resource as a file:

[pb82]

@AlexEndris it sounds like the Operator removes fields that are not in the spec and then ArgoCD applies them again. In that case, wouldn't it make sense to remove those fields (and use the 5m0s format for the other)?

[AlexEndris]

@pb82 Thank you for your reply. My point was that the yaml is just a copy & paste from what grafana itself exports in version 10.4.1. I assumed that those would be the same.

[theSuess]

The duration fields in the Kubernetes custom resource are using a different duration type than the Grafana export, that's why the saved resource shows a different to the Grafana export.

Does this cause any problems or is it just a cosmetic issue?

[AlexEndris]

@theSuess It actually causes ArgoCD to detect a difference and tries to constantly sync the state, overwriting the manifest over and over. We had to adjust the export manually to stop this.

[theSuess]

I see. I don't think there's anything we can do about this from an operator perspective (as the data representation is handled differently in k8s vs Grafana).

I'll close this issue and check with the alerting team to see if we can add a separate export format for an operator compatible CR into Grafana

[AlexEndris]

@theSuess Thank you for the consideration! That would definitely help reducing manually having to change this. ❤️