Update insight grafana docs

docs/en/docs/insight/quickstart/install/upgrade-note.md

@@ -7,18 +7,48 @@ Date: 2024-09-24
 
 This page provides some considerations for upgrading insight-server and insight-agent.
 
-## insight-agent
+## insight-server
 
-### Upgrade from v0.28.x (or lower) to v0.29.x
+### Upgrade from v0.39.x (or lower) to v0.40.x or higher
 
-Due to the upgrade of the Opentelemetry community operator chart version in v0.29.0, the supported values for `featureGates` in the values file have changed. Therefore, before upgrading, you need to set the value of `featureGates` to empty, as follows:
+In version v0.40.0, the Grafana Operator will be upgraded from v4 to v5, introducing significant CRD changes. 
+The upgrade process will be automatically completed by `helm upgrade`.
 
-```diff
--  --set opentelemetry-operator.manager.featureGates="+operator.autoinstrumentation.go,+operator.autoinstrumentation.multi-instrumentation,+operator.autoinstrumentation.nginx" \
-+  --set opentelemetry-operator.manager.featureGates=""
+If users need to perform a `helm rollback` to a previous version (e.g., from v0.40.0 to v0.39.0), 
+they must manually clean up the v5 Custom Resources (CRs) using the following commands:
+```shell
+kubectl delete grafanas.grafana.integreatly.org -n insight-system --selector operator.insight.io/managed-by=insight --ignore-not-found=true
+kubectl delete grafanadashboards.grafana.integreatly.org -n insight-system --selector operator.insight.io/managed-by=insight --ignore-not-found=true
+kubectl delete grafanadatasources.grafana.integreatly.org -n insight-system --selector operator.insight.io/managed-by=insight --ignore-not-found=true
 ```
+These commands only clean up CRs in the `insight-system` namespace. CRs in other namespaces can be removed using the same approach.
 
-## insight-server
+The Grafana Deployment in v0.40.x will include a [dashboard-discover](https://github.com/openinsight-proj/dashboard-discover) sidecar, 
+which is used to write specific GrafanaDashboard(v4) resources and ConfigMaps from the existing environment into the directory 
+specified by the Grafana dashboard provider (/var/lib/grafana/plugins/dashboards).
+
+
+The architecture diagram of dashboard-discover is as follows:
+
+![img](../../image/dashboard-discover.jpg)
+
+The specific rules are as follows:
+
+1. GrafanaDashboard(v4)
+
+   The dashboard-discover sidecar will watch for GrafanaDashboard(v4) resources across all namespaces in the cluster that 
+   have the label: `operator.insight.io/managed-by=insight`, and write their JSON content into the `/var/lib/grafana/plugins/dashboards`
+   directory of the Grafana container.
+
+
+2. ConfigMap
+
+   The dashboard-discover sidecar will watch for ConfigMaps across all namespaces in the cluster that have the labels: 
+   `operator.insight.io/managed-by=insight,operator.insight.io/dashboard=true`, and write all content under their data field 
+   into the `/var/lib/grafana/plugins/dashboards` directory of the Grafana container.
+
+> If you need to store JSON files in a specific folder, you can add the following label to the corresponding resource: 
+> `operator.insight.io/dashboard-folder=your-folder`
 
 ### Upgrade from v0.26.x (or lower) to v0.27.x or higher
 
@@ -86,6 +116,15 @@ kubectl apply --server-side -f https://raw.githubusercontent.com/VictoriaMetrics
 
 ## insight-agent
 
+### Upgrade from v0.28.x (or lower) to v0.29.x
+
+Due to the upgrade of the Opentelemetry community operator chart version in v0.29.0, the supported values for `featureGates` in the values file have changed. Therefore, before upgrading, you need to set the value of `featureGates` to empty, as follows:
+
+```diff
+-  --set opentelemetry-operator.manager.featureGates="+operator.autoinstrumentation.go,+operator.autoinstrumentation.multi-instrumentation,+operator.autoinstrumentation.nginx" \
++  --set opentelemetry-operator.manager.featureGates=""
+```
+
 ### Upgrade from v0.23.x (or lower) to v0.24.x
 
 In v0.24.x, CRDs have been added to the `OTEL operator chart`. However,

docs/en/docs/insight/user-guide/dashboard/import-dashboard.md

@@ -6,25 +6,29 @@ By using Grafana CRD, you can incorporate the management and deployment of dashb
 
 1. Log in to the DCE 5.0 platform and go to __Container Management__ . Select the __kpanda-global-cluster__ from the cluster list.
 
-2. Choose __Custom Resources__ from the left navigation bar. Look for the __grafanadashboards.integreatly.org__ 
+2. Choose __Custom Resources__ from the left navigation bar. Look for the __grafana.integreatly.org__ 
    file in the list and click it to view the details.
 
 3. Click __YAML Create__ and use the following template. Replace the dashboard JSON in the __Json__ field.
 
-    - __namespace__ : Specify the target namespace.
-    - __name__ : Provide a name for the dashboard.
-    - __label__ : Mandatory. Set the label as __operator.insight.io/managed-by: insight__ .
+   - __namespace__ ：填写目标命名空间；
+   - __name__ ：填写仪表盘的名称。
+   - __spec.allowCrossNamespaceImport__ ：必填，值为 __true__ 。
+   - __spec.instanceSelector.matchLabels__ ：必填，__operator.insight.io/managed-by: insight__ 。
 
     ```yaml
-    apiVersion: integreatly.org/v1alpha1
+    apiVersion: grafana.integreatly.org/v1beta1
     kind: GrafanaDashboard
     metadata:
       labels:
         app: insight-grafana-operator
-        operator.insight.io/managed-by: insight
       name: sample-dashboard
       namespace: insight-system
     spec:
+      allowCrossNamespaceImport: true
+      instanceSelector:
+        matchLabels:
+          operator.insight.io/managed-by: insight
       json: >
         {
           "id": null,

docs/zh/docs/insight/quickstart/install/upgrade-note.md

@@ -2,18 +2,70 @@
 
 本页介绍一些升级 insight-server 和 insight-agent 的注意事项。
 
-## insight-agent
+## insight-server
 
-### 从 v0.28.x（或更低版本）升级到 v0.29.x
+###  从 v0.39.x（或更低版本）升级到 v0.40.x 或更高版本
 
-由于 v0.29.0 升级了 Opentelemetry 社区的 operator chart 版本，values 中的 featureGates 的支持的值有所变化，因此，在 upgrade 之前，需要将 `featureGates` 的值设置为空, 即：
+在 v0.40.0 版本中，grafana operator 会从 v4 升级到 v5 并会引入大量的 CRD 变更。 升级过程会由 `helm upgrade` 自动完成。
+如果客户需要 `helm rollback` 回到老版本（比如，v0.40.0 -> v0.39.0）需要是手动清理 v5 的 CR:
 
-```diff
--  --set opentelemetry-operator.manager.featureGates="+operator.autoinstrumentation.go,+operator.autoinstrumentation.multi-instrumentation,+operator.autoinstrumentation.nginx" \
-+  --set opentelemetry-operator.manager.featureGates=""
+```shell
+kubectl delete grafanas.grafana.integreatly.org -n insight-system --selector operator.insight.io/managed-by=insight --ignore-not-found=true
+kubectl delete grafanadashboards.grafana.integreatly.org -n insight-system --selector operator.insight.io/managed-by=insight --ignore-not-found=true
+kubectl delete grafanadatasources.grafana.integreatly.org -n insight-system --selector operator.insight.io/managed-by=insight --ignore-not-found=true
 ```
 
-## insight-server
+这部分命令只清理了 insight-system 命名空间下的 CR, 其他命名空间可以用相同的方式处理。
+
+在 v0.40.x 的 grafana deployment 会添加一个 [dashboard-discover](https://github.com/openinsight-proj/dashboard-discover) sidecar
+用于将现有环境中特定的 GrafanaDashboard(v4)，ConfigMap 写入到 grafana dashboard provider 指定目录(/var/lib/grafana/plugins/dashboards)中。
+
+架构图如下：
+
+![img](../../images/upgrade-note01.jpg)
+
+具体规则如下：
+
+1. GrafanaDashboard(v4)
+
+   dashboard-discover sidecar 会 watch 集群中所用命名空间中带有：`operator.insight.io/managed-by=insight` label 的 GrafanaDashboard(v4)
+并将其 Json 内容写入到 Grafana 容器的 `/var/lib/grafana/plugins/dashboards` 中。
+
+
+2. ConfigMap
+
+   dashboard-discover sidecar 会 watch 集群中所用命名空间中带有：`operator.insight.io/managed-by=insight,operator.insight.io/dashboard=true` label 的 ConfigMap
+并将其 data 下的所有内容写入到 Grafana 容器的 `/var/lib/grafana/plugins/dashboards` 中。
+
+> 如果需要将 Json 文件存入特定文件夹中，可以在对应资源 label 中添加：`operator.insight.io/dashboard-folder=you-folder`
+
+在 v0.40.0 版本中， grafana 从 9.3.14 升级到 12.1.3。grafana 12.1.3 已完全移除对 AngularJS 的支持，转而优先支持 React。具体可以看[社区说明](https://grafana.com/blog/2025/04/03/angularjs-support-will-be-removed-in-grafana-12-what-you-need-to-know)。
+
+insight 以及兄弟团队维护的仪表盘已经自动迁移，开箱即用。对于客户自己维护的仪表盘，grafana 12.1.3 对部分核心预置（pre-installed）的
+AngularJS 面板提供自动迁移支持。在客户首次在 Grafana 12.1.3 打开自定义仪表盘后 grafana 会自动触发迁移, 迁移完成后，务必点击仪表盘的「Save」按钮保存变更，避免每次加载时重复迁移。
+
+若未完成 AngularJS 面板迁移，在 Grafana 12.1.3 后会出现以下问题：
+
+1. 依赖 AngularJS 的插件无法加载，在插件目录中不会显示已安装状态。
+2. 已配置的 AngularJS 数据源不会出现在数据源列表中。
+3. 仪表盘中原 AngularJS 面板会显示错误提示，如 `Error loading: plugin_name` 或 `Panel plugin not found: plugin_name`； 如下图：
+
+   <img src="../../images/upgrade-note02.png" alt="angularjs-deprecation-warning-panel" width="50%" />
+
+   可以手动替换类似的面板或插件。
+
+4. 仪表盘中原 Datasource 丢失，如 `Datasource XXX was not found`, 如下图：
+
+   <img src="../../images/upgrade-note03.png" alt="datasource-not-found" width="50%" />
+
+   可以手动创建一个 Datasource 变量:
+
+   <img src="../../images/upgrade-note04.png" alt="grafana-prometheus-ds-var" height="500" />
+
+   然后在面板中使用：
+
+   <img src="../../images/upgrade-note05.png" alt="grafana-use-prometheus-ds-var" width="30%" />
+
 
 ### 从 v0.26.x（或更低版本）升级到 v0.27.x 或更高版本
 
@@ -76,6 +128,15 @@ kubectl apply --server-side -f https://raw.githubusercontent.com/VictoriaMetrics
 
 ## insight-agent
 
+### 从 v0.28.x（或更低版本）升级到 v0.29.x
+
+由于 v0.29.0 升级了 Opentelemetry 社区的 operator chart 版本，values 中的 featureGates 的支持的值有所变化，因此，在 upgrade 之前，需要将 `featureGates` 的值设置为空, 即：
+
+```diff
+-  --set opentelemetry-operator.manager.featureGates="+operator.autoinstrumentation.go,+operator.autoinstrumentation.multi-instrumentation,+operator.autoinstrumentation.nginx" \
++  --set opentelemetry-operator.manager.featureGates=""
+```
+
 ### 从 v0.23.x（或更低版本）升级到 v0.24.x
 
 由于 0.24.x 版本中 `OTEL operator chart` 中新增了 CRD，但由于 Helm Upgrade 时并不会更新 CRD，因此，需要手动执行以下命令：

docs/zh/docs/insight/user-guide/dashboard/import-dashboard.md

@@ -6,26 +6,30 @@
 
 1. 登录 DCE 5.0 平台，进入 __容器管理__ ，在集群列表中选择 __kpanda-global-cluster__ 。
 
-2. 选择左侧导航栏的 __自定义资源__ ，在列表中查找 __grafanadashboards.integreatly.org__ 文件，进入详情。
+2. 选择左侧导航栏的 __自定义资源__ ，在列表中查找 __grafana.integreatly.org__ 文件，进入详情。
 
     ![导入仪表盘](https://docs.daocloud.io/daocloud-docs-images/docs/insight/images/importboard00.png){ width="1000"}
 
 3. 点击 __Yaml 创建__ ，使用以下模板，在 __Json__ 字段中替换仪表盘 JSON。
 
     - __namespace__ ：填写目标命名空间；
     - __name__ ：填写仪表盘的名称。
-    - __label__ ：必填， __operator.insight.io/managed-by: insight__ 。
+    - __spec.allowCrossNamespaceImport__ ：必填，值为 __true__ 。
+    - __spec.instanceSelector.matchLabels__ ：必填，__operator.insight.io/managed-by: insight__ 。
 
     ```yaml
-    apiVersion: integreatly.org/v1alpha1
+    apiVersion: grafana.integreatly.org/v1beta1
     kind: GrafanaDashboard
     metadata:
       labels:
         app: insight-grafana-operator
-        operator.insight.io/managed-by: insight
       name: sample-dashboard
       namespace: insight-system
     spec:
+      allowCrossNamespaceImport: true
+      instanceSelector:
+        matchLabels:
+          operator.insight.io/managed-by: insight
       json: >
         {
           "id": null,