Merge pull request 'feat/impl_installable_crd_prometheus' (#170) from feat/impl_installable_crd_prometheus into master

Reviewed-on: https://git.nationtech.io/NationTech/harmony/pulls/170

docs/doc-clone-and-restore-coreos.md

@@ -1,127 +0,0 @@
-## Working procedure to clone and restore CoreOS disk from OKD Cluster
-
-### **Step 1 - take a backup**
-```
-sudo dd if=/dev/old of=/dev/backup status=progress 
-```
-
-### **Step 2 - clone beginning of old disk to new**
-```
-sudo dd if=/dev/old of=/dev/backup status=progress count=1000Mib
-```
-
-### **Step 3 - verify and modify disk partitions**
-list disk partitions
-```
-sgdisk -p /dev/new
-```
-if new disk is smaller than old disk and there is space on the xfs partition of the old disk, modify partitions of new disk
-```
-gdisk /dev/new
-```
-inside of gdisk commands
-```
--v -> verify table
--p -> print table
--d -> select partition to delete partition
--n -> recreate partition with same partition number as deleted partition
-```
-For end sector, either specify the new end or just press Enter for maximum available 
-When asked about partition type, enter the same type code (it will show the old one) 
-```
-p - >to verify 
-w -> to write
-```
-make xfs file system for new partition <new4>
-```
-sudo mkfs.xfs -f /dev/new4
-```
-
-### **Step 4 - copy old PARTUUID **
-
-**careful here**
-get old patuuid:
-```
-sgdisk -i <partition_number> /dev/old_disk  # Note the "Partition unique GUID"
-```
-get labels
-```
-sgdisk -p /dev/old_disk  # Shows partition names in the table
-
-blkid /dev/old_disk*  # Shows PARTUUIDs and labels for all partitions
-```
-set it on new disk
-```
-sgdisk -u <partition_number>:<old_partuuid> /dev/sdc
-```
-partition name:
-```
-sgdisk -c <partition_number>:"<old_name>" /dev/sdc
-```
-verify all:
-```
-lsblk -o NAME,SIZE,PARTUUID,PARTLABEL /dev/old_disk
-```
-
-### **Step 5 - Mount disks and copy files from old to new disk**
-
-mount files before copy:
-
-```
-mkdir -p /mnt/new
-mkdir -p /mnt/old
-mount /dev/old4 /mnt/old
-mount /dev/new4 /mnt/new
-```
-copy:
-```
-rsync -aAXHv --numeric-ids /source/ /destination/
-```
-
-### **Step 6 - Set correct UUID for new partition 4**
-to set correct uuid for partition 4
-```
-blkid /dev/old4
-```
-```
-xfs_admin -U <old_uuid> /dev/new_partition
-```
-to set labels
-get it 
-```
-sgdisk -i 4 /dev/sda | grep "Partition name"
-```
-set it
-```
-sgdisk -c 4:"<label_name>" /dev/sdc
-
-or
-
-(check existing with xfs_admin -l /dev/old_partition)
-Use xfs_admin -L <label> /dev/new_partition 
-```
-
-### **Step 7 - Verify**
-
-verify everything:
-```
-sgdisk -p /dev/sda  # Old disk
-sgdisk -p /dev/sdc  # New disk
-```
-```
-lsblk -o NAME,SIZE,PARTUUID,PARTLABEL /dev/sda
-lsblk -o NAME,SIZE,PARTUUID,PARTLABEL /dev/sdc
-```
-```
-blkid /dev/sda* | grep UUID=
-blkid /dev/sdc* | grep UUID=
-```
-
-## **Step 8 - Unmount devices and Finalize job**
-unmount old devices
-```
-umount /mnt/new
-umount /mnt/old
-```
-
-shutdown swap disk and verify it worked

examples/try_rust_webapp/src/main.rs

@@ -3,7 +3,7 @@ use harmony::{
     modules::{
         application::{
             ApplicationScore, RustWebFramework, RustWebapp,
-            features::{PackagingDeployment, rhob_monitoring::Monitoring},
+            features::{Monitoring, PackagingDeployment},
         },
         monitoring::alert_channel::discord_alert_channel::DiscordWebhook,
     },

harmony/src/domain/topology/k8s.rs

@@ -3,7 +3,10 @@ use std::time::Duration;
 use derive_new::new;
 use k8s_openapi::{
     ClusterResourceScope, NamespaceResourceScope,
-    api::{apps::v1::Deployment, core::v1::Pod},
+    api::{
+        apps::v1::Deployment,
+        core::v1::{Pod, ServiceAccount},
+    },
     apimachinery::pkg::version::Info,
 };
 use kube::{
@@ -21,7 +24,7 @@ use kube::{
 };
 use log::{debug, error, info, trace};
 use serde::{Serialize, de::DeserializeOwned};
-use serde_json::{Value, json};
+use serde_json::json;
 use similar::TextDiff;
 use tokio::{io::AsyncReadExt, time::sleep};
 
@@ -57,6 +60,11 @@ impl K8sClient {
         })
     }
 
+    pub async fn service_account_api(&self, namespace: &str) -> Api<ServiceAccount> {
+        let api: Api<ServiceAccount> = Api::namespaced(self.client.clone(), namespace);
+        api
+    }
+
     pub async fn get_apiserver_version(&self) -> Result<Info, Error> {
         let client: Client = self.client.clone();
         let version_info: Info = client.apiserver_version().await?;

harmony/src/domain/topology/k8s_anywhere.rs

@@ -1,7 +1,12 @@
-use std::{process::Command, sync::Arc};
+use std::{collections::BTreeMap, process::Command, sync::Arc};
 
 use async_trait::async_trait;
-use kube::api::GroupVersionKind;
+use base64::{Engine, engine::general_purpose};
+use k8s_openapi::api::{
+    core::v1::Secret,
+    rbac::v1::{ClusterRoleBinding, RoleRef, Subject},
+};
+use kube::api::{DynamicObject, GroupVersionKind, ObjectMeta};
 use log::{debug, info, warn};
 use serde::Serialize;
 use tokio::sync::OnceCell;
@@ -12,14 +17,26 @@ use crate::{
     inventory::Inventory,
     modules::{
         k3d::K3DInstallationScore,
-        monitoring::kube_prometheus::crd::{
-            crd_alertmanager_config::CRDPrometheus,
-            prometheus_operator::prometheus_operator_helm_chart_score,
-            rhob_alertmanager_config::RHOBObservability,
+        k8s::ingress::{K8sIngressScore, PathType},
+        monitoring::{
+            grafana::{grafana::Grafana, helm::helm_grafana::grafana_helm_chart_score},
+            kube_prometheus::crd::{
+                crd_alertmanager_config::CRDPrometheus,
+                crd_grafana::{
+                    Grafana as GrafanaCRD, GrafanaCom, GrafanaDashboard,
+                    GrafanaDashboardDatasource, GrafanaDashboardSpec, GrafanaDatasource,
+                    GrafanaDatasourceConfig, GrafanaDatasourceJsonData,
+                    GrafanaDatasourceSecureJsonData, GrafanaDatasourceSpec, GrafanaSpec,
+                },
+                crd_prometheuses::LabelSelector,
+                prometheus_operator::prometheus_operator_helm_chart_score,
+                rhob_alertmanager_config::RHOBObservability,
+                service_monitor::ServiceMonitor,
+            },
         },
         prometheus::{
             k8s_prometheus_alerting_score::K8sPrometheusCRDAlertingScore,
-            prometheus::PrometheusApplicationMonitoring, rhob_alerting_score::RHOBAlertingScore,
+            prometheus::PrometheusMonitoring, rhob_alerting_score::RHOBAlertingScore,
         },
     },
     score::Score,
@@ -86,41 +103,172 @@ impl K8sclient for K8sAnywhereTopology {
 }
 
 #[async_trait]
-impl PrometheusApplicationMonitoring<CRDPrometheus> for K8sAnywhereTopology {
+impl Grafana for K8sAnywhereTopology {
+    async fn ensure_grafana_operator(
+        &self,
+        inventory: &Inventory,
+    ) -> Result<PreparationOutcome, PreparationError> {
+        debug!("ensure grafana operator");
+        let client = self.k8s_client().await.unwrap();
+        let grafana_gvk = GroupVersionKind {
+            group: "grafana.integreatly.org".to_string(),
+            version: "v1beta1".to_string(),
+            kind: "Grafana".to_string(),
+        };
+        let name = "grafanas.grafana.integreatly.org";
+        let ns = "grafana";
+
+        let grafana_crd = client
+            .get_resource_json_value(name, Some(ns), &grafana_gvk)
+            .await;
+        match grafana_crd {
+            Ok(_) => {
+                return Ok(PreparationOutcome::Success {
+                    details: "Found grafana CRDs in cluster".to_string(),
+                });
+            }
+
+            Err(_) => {
+                return self
+                    .install_grafana_operator(inventory, Some("grafana"))
+                    .await;
+            }
+        };
+    }
+    async fn install_grafana(&self) -> Result<PreparationOutcome, PreparationError> {
+        let ns = "grafana";
+
+        let mut label = BTreeMap::new();
+
+        label.insert("dashboards".to_string(), "grafana".to_string());
+
+        let label_selector = LabelSelector {
+            match_labels: label.clone(),
+            match_expressions: vec![],
+        };
+
+        let client = self.k8s_client().await?;
+
+        let grafana = self.build_grafana(ns, &label);
+
+        client.apply(&grafana, Some(ns)).await?;
+        //TODO change this to a ensure ready or something better than just a timeout
+        client
+            .wait_until_deployment_ready(
+                "grafana-grafana-deployment".to_string(),
+                Some("grafana"),
+                Some(30),
+            )
+            .await?;
+
+        let sa_name = "grafana-grafana-sa";
+        let token_secret_name = "grafana-sa-token-secret";
+
+        let sa_token_secret = self.build_sa_token_secret(token_secret_name, sa_name, ns);
+
+        client.apply(&sa_token_secret, Some(ns)).await?;
+        let secret_gvk = GroupVersionKind {
+            group: "".to_string(),
+            version: "v1".to_string(),
+            kind: "Secret".to_string(),
+        };
+
+        let secret = client
+            .get_resource_json_value(token_secret_name, Some(ns), &secret_gvk)
+            .await?;
+
+        let token = format!(
+            "Bearer {}",
+            self.extract_and_normalize_token(&secret).unwrap()
+        );
+
+        debug!("creating grafana clusterrole binding");
+
+        let clusterrolebinding =
+            self.build_cluster_rolebinding(sa_name, "cluster-monitoring-view", ns);
+
+        client.apply(&clusterrolebinding, Some(ns)).await?;
+
+        debug!("creating grafana datasource crd");
+
+        let thanos_url = format!(
+            "https://{}",
+            self.get_domain("thanos-querier-openshift-monitoring")
+                .await
+                .unwrap()
+        );
+
+        let thanos_openshift_datasource = self.build_grafana_datasource(
+            "thanos-openshift-monitoring",
+            ns,
+            &label_selector,
+            &thanos_url,
+            &token,
+        );
+
+        client.apply(&thanos_openshift_datasource, Some(ns)).await?;
+
+        debug!("creating grafana dashboard crd");
+        let dashboard = self.build_grafana_dashboard(ns, &label_selector);
+
+        client.apply(&dashboard, Some(ns)).await?;
+        debug!("creating grafana ingress");
+        let grafana_ingress = self.build_grafana_ingress(ns).await;
+
+        grafana_ingress
+            .interpret(&Inventory::empty(), self)
+            .await
+            .map_err(|e| PreparationError::new(e.to_string()))?;
+
+        Ok(PreparationOutcome::Success {
+            details: "Installed grafana composants".to_string(),
+        })
+    }
+}
+
+#[async_trait]
+impl PrometheusMonitoring<CRDPrometheus> for K8sAnywhereTopology {
     async fn install_prometheus(
         &self,
         sender: &CRDPrometheus,
-        inventory: &Inventory,
-        receivers: Option<Vec<Box<dyn AlertReceiver<CRDPrometheus>>>>,
+        _inventory: &Inventory,
+        _receivers: Option<Vec<Box<dyn AlertReceiver<CRDPrometheus>>>>,
+    ) -> Result<PreparationOutcome, PreparationError> {
+        let client = self.k8s_client().await?;
+
+        for monitor in sender.service_monitor.iter() {
+            client
+                .apply(monitor, Some(&sender.namespace))
+                .await
+                .map_err(|e| PreparationError::new(e.to_string()))?;
+        }
+        Ok(PreparationOutcome::Success {
+            details: "successfuly installed prometheus components".to_string(),
+        })
+    }
+
+    async fn ensure_prometheus_operator(
+        &self,
+        sender: &CRDPrometheus,
+        _inventory: &Inventory,
     ) -> Result<PreparationOutcome, PreparationError> {
         let po_result = self.ensure_prometheus_operator(sender).await?;
 
-        if po_result == PreparationOutcome::Noop {
-            debug!("Skipping Prometheus CR installation due to missing operator.");
-            return Ok(po_result);
-        }
-
-        let result = self
-            .get_k8s_prometheus_application_score(sender.clone(), receivers)
-            .await
-            .interpret(inventory, self)
-            .await;
-
-        match result {
-            Ok(outcome) => match outcome.status {
-                InterpretStatus::SUCCESS => Ok(PreparationOutcome::Success {
-                    details: outcome.message,
-                }),
-                InterpretStatus::NOOP => Ok(PreparationOutcome::Noop),
-                _ => Err(PreparationError::new(outcome.message)),
-            },
-            Err(err) => Err(PreparationError::new(err.to_string())),
+        match po_result {
+            PreparationOutcome::Success { details: _ } => {
+                debug!("Detected prometheus crds operator present in cluster.");
+                return Ok(po_result);
+            }
+            PreparationOutcome::Noop => {
+                debug!("Skipping Prometheus CR installation due to missing operator.");
+                return Ok(po_result);
+            }
         }
     }
 }
 
 #[async_trait]
-impl PrometheusApplicationMonitoring<RHOBObservability> for K8sAnywhereTopology {
+impl PrometheusMonitoring<RHOBObservability> for K8sAnywhereTopology {
     async fn install_prometheus(
         &self,
         sender: &RHOBObservability,
@@ -154,6 +302,14 @@ impl PrometheusApplicationMonitoring<RHOBObservability> for K8sAnywhereTopology
             Err(err) => Err(PreparationError::new(err.to_string())),
         }
     }
+
+    async fn ensure_prometheus_operator(
+        &self,
+        sender: &RHOBObservability,
+        inventory: &Inventory,
+    ) -> Result<PreparationOutcome, PreparationError> {
+        todo!()
+    }
 }
 
 impl Serialize for K8sAnywhereTopology {
@@ -215,6 +371,180 @@ impl K8sAnywhereTopology {
             .await
     }
 
+    fn extract_and_normalize_token(&self, secret: &DynamicObject) -> Option<String> {
+        let token_b64 = secret
+            .data
+            .get("token")
+            .or_else(|| secret.data.get("data").and_then(|d| d.get("token")))
+            .and_then(|v| v.as_str())?;
+
+        let bytes = general_purpose::STANDARD.decode(token_b64).ok()?;
+
+        let s = String::from_utf8(bytes).ok()?;
+
+        let cleaned = s
+            .trim_matches(|c: char| c.is_whitespace() || c == '\0')
+            .to_string();
+        Some(cleaned)
+    }
+
+    pub fn build_cluster_rolebinding(
+        &self,
+        service_account_name: &str,
+        clusterrole_name: &str,
+        ns: &str,
+    ) -> ClusterRoleBinding {
+        ClusterRoleBinding {
+            metadata: ObjectMeta {
+                name: Some(format!("{}-view-binding", service_account_name)),
+                ..Default::default()
+            },
+            role_ref: RoleRef {
+                api_group: "rbac.authorization.k8s.io".into(),
+                kind: "ClusterRole".into(),
+                name: clusterrole_name.into(),
+            },
+            subjects: Some(vec![Subject {
+                kind: "ServiceAccount".into(),
+                name: service_account_name.into(),
+                namespace: Some(ns.into()),
+                ..Default::default()
+            }]),
+        }
+    }
+
+    pub fn build_sa_token_secret(
+        &self,
+        secret_name: &str,
+        service_account_name: &str,
+        ns: &str,
+    ) -> Secret {
+        let mut annotations = BTreeMap::new();
+        annotations.insert(
+            "kubernetes.io/service-account.name".to_string(),
+            service_account_name.to_string(),
+        );
+
+        Secret {
+            metadata: ObjectMeta {
+                name: Some(secret_name.into()),
+                namespace: Some(ns.into()),
+                annotations: Some(annotations),
+                ..Default::default()
+            },
+            type_: Some("kubernetes.io/service-account-token".to_string()),
+            ..Default::default()
+        }
+    }
+
+    fn build_grafana_datasource(
+        &self,
+        name: &str,
+        ns: &str,
+        label_selector: &LabelSelector,
+        url: &str,
+        token: &str,
+    ) -> GrafanaDatasource {
+        let mut json_data = BTreeMap::new();
+        json_data.insert("timeInterval".to_string(), "5s".to_string());
+
+        GrafanaDatasource {
+            metadata: ObjectMeta {
+                name: Some(name.to_string()),
+                namespace: Some(ns.to_string()),
+                ..Default::default()
+            },
+            spec: GrafanaDatasourceSpec {
+                instance_selector: label_selector.clone(),
+                allow_cross_namespace_import: Some(true),
+                values_from: None,
+                datasource: GrafanaDatasourceConfig {
+                    access: "proxy".to_string(),
+                    name: name.to_string(),
+                    r#type: "prometheus".to_string(),
+                    url: url.to_string(),
+                    database: None,
+                    json_data: Some(GrafanaDatasourceJsonData {
+                        time_interval: Some("60s".to_string()),
+                        http_header_name1: Some("Authorization".to_string()),
+                        tls_skip_verify: Some(true),
+                        oauth_pass_thru: Some(true),
+                    }),
+                    secure_json_data: Some(GrafanaDatasourceSecureJsonData {
+                        http_header_value1: Some(format!("Bearer {token}")),
+                    }),
+                    is_default: Some(false),
+                    editable: Some(true),
+                },
+            },
+        }
+    }
+
+    fn build_grafana_dashboard(
+        &self,
+        ns: &str,
+        label_selector: &LabelSelector,
+    ) -> GrafanaDashboard {
+        let graf_dashboard = GrafanaDashboard {
+            metadata: ObjectMeta {
+                name: Some(format!("grafana-dashboard-{}", ns)),
+                namespace: Some(ns.to_string()),
+                ..Default::default()
+            },
+            spec: GrafanaDashboardSpec {
+                resync_period: Some("30s".to_string()),
+                instance_selector: label_selector.clone(),
+                datasources: Some(vec![GrafanaDashboardDatasource {
+                    input_name: "DS_PROMETHEUS".to_string(),
+                    datasource_name: "thanos-openshift-monitoring".to_string(),
+                }]),
+                json: None,
+                grafana_com: Some(GrafanaCom {
+                    id: 17406,
+                    revision: None,
+                }),
+            },
+        };
+        graf_dashboard
+    }
+
+    fn build_grafana(&self, ns: &str, labels: &BTreeMap<String, String>) -> GrafanaCRD {
+        let grafana = GrafanaCRD {
+            metadata: ObjectMeta {
+                name: Some(format!("grafana-{}", ns)),
+                namespace: Some(ns.to_string()),
+                labels: Some(labels.clone()),
+                ..Default::default()
+            },
+            spec: GrafanaSpec {
+                config: None,
+                admin_user: None,
+                admin_password: None,
+                ingress: None,
+                persistence: None,
+                resources: None,
+            },
+        };
+        grafana
+    }
+
+    async fn build_grafana_ingress(&self, ns: &str) -> K8sIngressScore {
+        let domain = self.get_domain(&format!("grafana-{}", ns)).await.unwrap();
+        let name = format!("{}-grafana", ns);
+        let backend_service = format!("grafana-{}-service", ns);
+
+        K8sIngressScore {
+            name: fqdn::fqdn!(&name),
+            host: fqdn::fqdn!(&domain),
+            backend_service: fqdn::fqdn!(&backend_service),
+            port: 3000,
+            path: Some("/".to_string()),
+            path_type: Some(PathType::Prefix),
+            namespace: Some(fqdn::fqdn!(&ns)),
+            ingress_class_name: Some("openshift-default".to_string()),
+        }
+    }
+
     async fn get_cluster_observability_operator_prometheus_application_score(
         &self,
         sender: RHOBObservability,
@@ -232,13 +562,14 @@ impl K8sAnywhereTopology {
         &self,
         sender: CRDPrometheus,
         receivers: Option<Vec<Box<dyn AlertReceiver<CRDPrometheus>>>>,
+        service_monitors: Option<Vec<ServiceMonitor>>,
     ) -> K8sPrometheusCRDAlertingScore {
-        K8sPrometheusCRDAlertingScore {
+        return K8sPrometheusCRDAlertingScore {
             sender,
             receivers: receivers.unwrap_or_default(),
-            service_monitors: vec![],
+            service_monitors: service_monitors.unwrap_or_default(),
             prometheus_rules: vec![],
-        }
+        };
     }
 
     async fn openshift_ingress_operator_available(&self) -> Result<(), PreparationError> {
@@ -506,6 +837,30 @@ impl K8sAnywhereTopology {
             details: "prometheus operator present in cluster".into(),
         })
     }
+
+    async fn install_grafana_operator(
+        &self,
+        inventory: &Inventory,
+        ns: Option<&str>,
+    ) -> Result<PreparationOutcome, PreparationError> {
+        let namespace = ns.unwrap_or("grafana");
+        info!("installing grafana operator in ns {namespace}");
+        let tenant = self.get_k8s_tenant_manager()?.get_tenant_config().await;
+        let mut namespace_scope = false;
+        if tenant.is_some() {
+            namespace_scope = true;
+        }
+        let _grafana_operator_score = grafana_helm_chart_score(namespace, namespace_scope)
+            .interpret(inventory, self)
+            .await
+            .map_err(|e| PreparationError::new(e.to_string()));
+        Ok(PreparationOutcome::Success {
+            details: format!(
+                "Successfully installed grafana operator in ns {}",
+                ns.unwrap()
+            ),
+        })
+    }
 }
 
 #[derive(Clone, Debug)]

harmony/src/domain/topology/oberservability/monitoring.rs

@@ -31,6 +31,7 @@ impl<S: AlertSender + Installable<T>, T: Topology> Interpret<T> for AlertingInte
         inventory: &Inventory,
         topology: &T,
     ) -> Result<Outcome, InterpretError> {
+        debug!("hit sender configure for AlertingInterpret");
         self.sender.configure(inventory, topology).await?;
         for receiver in self.receivers.iter() {
             receiver.install(&self.sender).await?;
@@ -86,4 +87,5 @@ pub trait AlertRule<S: AlertSender>: std::fmt::Debug + Send + Sync {
 #[async_trait]
 pub trait ScrapeTarget<S: AlertSender>: std::fmt::Debug + Send + Sync {
     async fn install(&self, sender: &S) -> Result<Outcome, InterpretError>;
+    fn clone_box(&self) -> Box<dyn ScrapeTarget<S>>;
 }

harmony/src/modules/application/features/monitoring.rs

@@ -2,7 +2,11 @@ use crate::modules::application::{
     Application, ApplicationFeature, InstallationError, InstallationOutcome,
 };
 use crate::modules::monitoring::application_monitoring::application_monitoring_score::ApplicationMonitoringScore;
+use crate::modules::monitoring::grafana::grafana::Grafana;
 use crate::modules::monitoring::kube_prometheus::crd::crd_alertmanager_config::CRDPrometheus;
+use crate::modules::monitoring::kube_prometheus::crd::service_monitor::{
+    ServiceMonitor, ServiceMonitorSpec,
+};
 use crate::topology::MultiTargetTopology;
 use crate::topology::ingress::Ingress;
 use crate::{
@@ -14,7 +18,7 @@ use crate::{
     topology::{HelmCommand, K8sclient, Topology, tenant::TenantManager},
 };
 use crate::{
-    modules::prometheus::prometheus::PrometheusApplicationMonitoring,
+    modules::prometheus::prometheus::PrometheusMonitoring,
     topology::oberservability::monitoring::AlertReceiver,
 };
 use async_trait::async_trait;
@@ -22,6 +26,7 @@ use base64::{Engine as _, engine::general_purpose};
 use harmony_secret::SecretManager;
 use harmony_secret_derive::Secret;
 use harmony_types::net::Url;
+use kube::api::ObjectMeta;
 use log::{debug, info};
 use serde::{Deserialize, Serialize};
 use std::sync::Arc;
@@ -40,7 +45,8 @@ impl<
         + TenantManager
         + K8sclient
         + MultiTargetTopology
-        + PrometheusApplicationMonitoring<CRDPrometheus>
+        + PrometheusMonitoring<CRDPrometheus>
+        + Grafana
         + Ingress
         + std::fmt::Debug,
 > ApplicationFeature<T> for Monitoring
@@ -57,10 +63,20 @@ impl<
             .unwrap_or_else(|| self.application.name());
         let domain = topology.get_domain("ntfy").await.unwrap();
 
+        let app_service_monitor = ServiceMonitor {
+            metadata: ObjectMeta {
+                name: Some(self.application.name()),
+                namespace: Some(namespace.clone()),
+                ..Default::default()
+            },
+            spec: ServiceMonitorSpec::default(),
+        };
+
         let mut alerting_score = ApplicationMonitoringScore {
             sender: CRDPrometheus {
                 namespace: namespace.clone(),
                 client: topology.k8s_client().await.unwrap(),
+                service_monitor: vec![app_service_monitor],
             },
             application: self.application.clone(),
             receivers: self.alert_receiver.clone(),

harmony/src/modules/application/features/rhob_monitoring.rs

@@ -18,7 +18,7 @@ use crate::{
     topology::{HelmCommand, K8sclient, Topology, tenant::TenantManager},
 };
 use crate::{
-    modules::prometheus::prometheus::PrometheusApplicationMonitoring,
+    modules::prometheus::prometheus::PrometheusMonitoring,
     topology::oberservability::monitoring::AlertReceiver,
 };
 use async_trait::async_trait;
@@ -42,7 +42,7 @@ impl<
         + MultiTargetTopology
         + Ingress
         + std::fmt::Debug
-        + PrometheusApplicationMonitoring<RHOBObservability>,
+        + PrometheusMonitoring<RHOBObservability>,
 > ApplicationFeature<T> for Monitoring
 {
     async fn ensure_installed(

harmony/src/modules/monitoring/application_monitoring/application_monitoring_score.rs

@@ -1,21 +1,23 @@
 use std::sync::Arc;
 
-use async_trait::async_trait;
+use log::debug;
 use serde::Serialize;
 
 use crate::{
-    data::Version,
-    interpret::{Interpret, InterpretError, InterpretName, InterpretStatus, Outcome},
-    inventory::Inventory,
+    interpret::Interpret,
     modules::{
         application::Application,
-        monitoring::kube_prometheus::crd::crd_alertmanager_config::CRDPrometheus,
-        prometheus::prometheus::PrometheusApplicationMonitoring,
+        monitoring::{
+            grafana::grafana::Grafana, kube_prometheus::crd::crd_alertmanager_config::CRDPrometheus,
+        },
+        prometheus::prometheus::PrometheusMonitoring,
     },
     score::Score,
-    topology::{PreparationOutcome, Topology, oberservability::monitoring::AlertReceiver},
+    topology::{
+        K8sclient, Topology,
+        oberservability::monitoring::{AlertReceiver, AlertingInterpret, ScrapeTarget},
+    },
 };
-use harmony_types::id::Id;
 
 #[derive(Debug, Clone, Serialize)]
 pub struct ApplicationMonitoringScore {
@@ -24,12 +26,16 @@ pub struct ApplicationMonitoringScore {
     pub receivers: Vec<Box<dyn AlertReceiver<CRDPrometheus>>>,
 }
 
-impl<T: Topology + PrometheusApplicationMonitoring<CRDPrometheus>> Score<T>
+impl<T: Topology + PrometheusMonitoring<CRDPrometheus> + K8sclient + Grafana> Score<T>
     for ApplicationMonitoringScore
 {
     fn create_interpret(&self) -> Box<dyn Interpret<T>> {
-        Box::new(ApplicationMonitoringInterpret {
-            score: self.clone(),
+        debug!("creating alerting interpret");
+        Box::new(AlertingInterpret {
+            sender: self.sender.clone(),
+            receivers: self.receivers.clone(),
+            rules: vec![],
+            scrape_targets: None,
         })
     }
 
@@ -40,55 +46,3 @@ impl<T: Topology + PrometheusApplicationMonitoring<CRDPrometheus>> Score<T>
         )
     }
 }
-
-#[derive(Debug)]
-pub struct ApplicationMonitoringInterpret {
-    score: ApplicationMonitoringScore,
-}
-
-#[async_trait]
-impl<T: Topology + PrometheusApplicationMonitoring<CRDPrometheus>> Interpret<T>
-    for ApplicationMonitoringInterpret
-{
-    async fn execute(
-        &self,
-        inventory: &Inventory,
-        topology: &T,
-    ) -> Result<Outcome, InterpretError> {
-        let result = topology
-            .install_prometheus(
-                &self.score.sender,
-                inventory,
-                Some(self.score.receivers.clone()),
-            )
-            .await;
-
-        match result {
-            Ok(outcome) => match outcome {
-                PreparationOutcome::Success { details: _ } => {
-                    Ok(Outcome::success("Prometheus installed".into()))
-                }
-                PreparationOutcome::Noop => {
-                    Ok(Outcome::noop("Prometheus installation skipped".into()))
-                }
-            },
-            Err(err) => Err(InterpretError::from(err)),
-        }
-    }
-
-    fn get_name(&self) -> InterpretName {
-        InterpretName::ApplicationMonitoring
-    }
-
-    fn get_version(&self) -> Version {
-        todo!()
-    }
-
-    fn get_status(&self) -> InterpretStatus {
-        todo!()
-    }
-
-    fn get_children(&self) -> Vec<Id> {
-        todo!()
-    }
-}

harmony/src/modules/monitoring/application_monitoring/rhobs_application_monitoring_score.rs

@@ -12,7 +12,7 @@ use crate::{
         monitoring::kube_prometheus::crd::{
             crd_alertmanager_config::CRDPrometheus, rhob_alertmanager_config::RHOBObservability,
         },
-        prometheus::prometheus::PrometheusApplicationMonitoring,
+        prometheus::prometheus::PrometheusMonitoring,
     },
     score::Score,
     topology::{PreparationOutcome, Topology, oberservability::monitoring::AlertReceiver},
@@ -26,7 +26,7 @@ pub struct ApplicationRHOBMonitoringScore {
     pub receivers: Vec<Box<dyn AlertReceiver<RHOBObservability>>>,
 }
 
-impl<T: Topology + PrometheusApplicationMonitoring<RHOBObservability>> Score<T>
+impl<T: Topology + PrometheusMonitoring<RHOBObservability>> Score<T>
     for ApplicationRHOBMonitoringScore
 {
     fn create_interpret(&self) -> Box<dyn Interpret<T>> {
@@ -49,7 +49,7 @@ pub struct ApplicationRHOBMonitoringInterpret {
 }
 
 #[async_trait]
-impl<T: Topology + PrometheusApplicationMonitoring<RHOBObservability>> Interpret<T>
+impl<T: Topology + PrometheusMonitoring<RHOBObservability>> Interpret<T>
     for ApplicationRHOBMonitoringInterpret
 {
     async fn execute(

harmony/src/modules/monitoring/grafana/grafana.rs

@@ -0,0 +1,17 @@
+use async_trait::async_trait;
+use k8s_openapi::Resource;
+
+use crate::{
+    inventory::Inventory,
+    topology::{PreparationError, PreparationOutcome},
+};
+
+#[async_trait]
+pub trait Grafana {
+    async fn ensure_grafana_operator(
+        &self,
+        inventory: &Inventory,
+    ) -> Result<PreparationOutcome, PreparationError>;
+
+    async fn install_grafana(&self) -> Result<PreparationOutcome, PreparationError>;
+}

harmony/src/modules/monitoring/grafana/helm/helm_grafana.rs

@@ -1,27 +1,28 @@
+use harmony_macros::hurl;
 use non_blank_string_rs::NonBlankString;
-use std::str::FromStr;
+use std::{collections::HashMap, str::FromStr};
 
-use crate::modules::helm::chart::HelmChartScore;
-
-pub fn grafana_helm_chart_score(ns: &str) -> HelmChartScore {
-    let values = r#"
-rbac:
-  namespaced: true
-sidecar:
-  dashboards:
-    enabled: true
-        "#
-    .to_string();
+use crate::modules::helm::chart::{HelmChartScore, HelmRepository};
 
+pub fn grafana_helm_chart_score(ns: &str, namespace_scope: bool) -> HelmChartScore {
+    let mut values_overrides = HashMap::new();
+    values_overrides.insert(
+        NonBlankString::from_str("namespaceScope").unwrap(),
+        namespace_scope.to_string(),
+    );
     HelmChartScore {
         namespace: Some(NonBlankString::from_str(ns).unwrap()),
-        release_name: NonBlankString::from_str("grafana").unwrap(),
-        chart_name: NonBlankString::from_str("oci://ghcr.io/grafana/helm-charts/grafana").unwrap(),
+        release_name: NonBlankString::from_str("grafana-operator").unwrap(),
+        chart_name: NonBlankString::from_str("grafana/grafana-operator").unwrap(),
         chart_version: None,
-        values_overrides: None,
-        values_yaml: Some(values.to_string()),
+        values_overrides: Some(values_overrides),
+        values_yaml: None,
         create_namespace: true,
         install_only: true,
-        repository: None,
+        repository: Some(HelmRepository::new(
+            "grafana".to_string(),
+            hurl!("https://grafana.github.io/helm-charts"),
+            true,
+        )),
     }
 }

harmony/src/modules/monitoring/grafana/mod.rs

@@ -1 +1,2 @@
+pub mod grafana;
 pub mod helm;

harmony/src/modules/monitoring/kube_prometheus/crd/crd_alertmanager_config.rs

@@ -1,12 +1,25 @@
 use std::sync::Arc;
 
+use async_trait::async_trait;
 use kube::CustomResource;
 use schemars::JsonSchema;
 use serde::{Deserialize, Serialize};
 
-use crate::topology::{
-    k8s::K8sClient,
-    oberservability::monitoring::{AlertReceiver, AlertSender},
+use crate::{
+    interpret::{InterpretError, Outcome},
+    inventory::Inventory,
+    modules::{
+        monitoring::{
+            grafana::grafana::Grafana, kube_prometheus::crd::service_monitor::ServiceMonitor,
+        },
+        prometheus::prometheus::PrometheusMonitoring,
+    },
+    topology::{
+        K8sclient, Topology,
+        installable::Installable,
+        k8s::K8sClient,
+        oberservability::monitoring::{AlertReceiver, AlertSender, ScrapeTarget},
+    },
 };
 
 #[derive(CustomResource, Serialize, Deserialize, Debug, Clone, JsonSchema)]
@@ -26,6 +39,7 @@ pub struct AlertmanagerConfigSpec {
 pub struct CRDPrometheus {
     pub namespace: String,
     pub client: Arc<K8sClient>,
+    pub service_monitor: Vec<ServiceMonitor>,
 }
 
 impl AlertSender for CRDPrometheus {
@@ -40,6 +54,12 @@ impl Clone for Box<dyn AlertReceiver<CRDPrometheus>> {
     }
 }
 
+impl Clone for Box<dyn ScrapeTarget<CRDPrometheus>> {
+    fn clone(&self) -> Self {
+        self.clone_box()
+    }
+}
+
 impl Serialize for Box<dyn AlertReceiver<CRDPrometheus>> {
     fn serialize<S>(&self, _serializer: S) -> Result<S::Ok, S::Error>
     where
@@ -48,3 +68,24 @@ impl Serialize for Box<dyn AlertReceiver<CRDPrometheus>> {
         todo!()
     }
 }
+
+#[async_trait]
+impl<T: Topology + K8sclient + PrometheusMonitoring<CRDPrometheus> + Grafana> Installable<T>
+    for CRDPrometheus
+{
+    async fn configure(&self, inventory: &Inventory, topology: &T) -> Result<(), InterpretError> {
+        topology.ensure_grafana_operator(inventory).await?;
+        topology.ensure_prometheus_operator(self, inventory).await?;
+        Ok(())
+    }
+
+    async fn ensure_installed(
+        &self,
+        inventory: &Inventory,
+        topology: &T,
+    ) -> Result<(), InterpretError> {
+        topology.install_grafana().await?;
+        topology.install_prometheus(&self, inventory, None).await?;
+        Ok(())
+    }
+}

harmony/src/modules/monitoring/kube_prometheus/crd/crd_grafana.rs

@@ -103,9 +103,34 @@ pub struct GrafanaDashboardSpec {
     #[serde(default, skip_serializing_if = "Option::is_none")]
     pub resync_period: Option<String>,
 
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub datasources: Option<Vec<GrafanaDashboardDatasource>>,
+
     pub instance_selector: LabelSelector,
 
-    pub json: String,
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub json: Option<String>,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub grafana_com: Option<GrafanaCom>,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaDashboardDatasource {
+    pub input_name: String,
+    pub datasource_name: String,
+}
+
+// ------------------------------------------------------------------------------------------------
+
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaCom {
+    pub id: u32,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub revision: Option<u32>,
 }
 
 // ------------------------------------------------------------------------------------------------
@@ -126,20 +151,79 @@ pub struct GrafanaDatasourceSpec {
     pub allow_cross_namespace_import: Option<bool>,
 
     pub datasource: GrafanaDatasourceConfig,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub values_from: Option<Vec<GrafanaValueFrom>>,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaValueFrom {
+    pub target_path: String,
+    pub value_from: GrafanaValueSource,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaValueSource {
+    pub secret_key_ref: GrafanaSecretKeyRef,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaSecretKeyRef {
+    pub name: String,
+    pub key: String,
 }
 
 #[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
 #[serde(rename_all = "camelCase")]
 pub struct GrafanaDatasourceConfig {
     pub access: String,
-    pub database: Option<String>,
     #[serde(default, skip_serializing_if = "Option::is_none")]
-    pub json_data: Option<BTreeMap<String, String>>,
+    pub database: Option<String>,
     pub name: String,
     pub r#type: String,
     pub url: String,
+    /// Represents jsonData in the GrafanaDatasource spec
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub json_data: Option<GrafanaDatasourceJsonData>,
+
+    /// Represents secureJsonData (secrets)
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub secure_json_data: Option<GrafanaDatasourceSecureJsonData>,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub is_default: Option<bool>,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub editable: Option<bool>,
 }
 
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaDatasourceJsonData {
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub time_interval: Option<String>,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub http_header_name1: Option<String>,
+
+    /// Disable TLS skip verification (false = verify)
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub tls_skip_verify: Option<bool>,
+
+    /// Auth type - set to "forward" for OpenShift OAuth identity
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub oauth_pass_thru: Option<bool>,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct GrafanaDatasourceSecureJsonData {
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub http_header_value1: Option<String>,
+}
 // ------------------------------------------------------------------------------------------------
 
 #[derive(Serialize, Deserialize, Debug, Clone, JsonSchema, Default)]

harmony/src/modules/monitoring/prometheus/prometheus.rs

@@ -114,7 +114,7 @@ impl Prometheus {
         };
 
         if let Some(ns) = namespace.as_deref() {
-            grafana_helm_chart_score(ns)
+            grafana_helm_chart_score(ns, false)
                 .interpret(inventory, topology)
                 .await
         } else {

harmony/src/modules/monitoring/scrape_target/server.rs

@@ -73,4 +73,8 @@ impl ScrapeTarget<CRDPrometheus> for Server {
             self.name.clone()
         )))
     }
+
+    fn clone_box(&self) -> Box<dyn ScrapeTarget<CRDPrometheus>> {
+        Box::new(self.clone())
+    }
 }

harmony/src/modules/prometheus/k8s_prometheus_alerting_score.rs

@@ -12,7 +12,8 @@ use crate::modules::monitoring::kube_prometheus::crd::crd_alertmanager_config::C
 use crate::modules::monitoring::kube_prometheus::crd::crd_default_rules::build_default_application_rules;
 use crate::modules::monitoring::kube_prometheus::crd::crd_grafana::{
     Grafana, GrafanaDashboard, GrafanaDashboardSpec, GrafanaDatasource, GrafanaDatasourceConfig,
-    GrafanaDatasourceSpec, GrafanaSpec,
+    GrafanaDatasourceJsonData, GrafanaDatasourceSpec, GrafanaSecretKeyRef, GrafanaSpec,
+    GrafanaValueFrom, GrafanaValueSource,
 };
 use crate::modules::monitoring::kube_prometheus::crd::crd_prometheus_rules::{
     PrometheusRule, PrometheusRuleSpec, RuleGroup,
@@ -39,7 +40,7 @@ use crate::{
 };
 use harmony_types::id::Id;
 
-use super::prometheus::PrometheusApplicationMonitoring;
+use super::prometheus::PrometheusMonitoring;
 
 #[derive(Clone, Debug, Serialize)]
 pub struct K8sPrometheusCRDAlertingScore {
@@ -49,7 +50,7 @@ pub struct K8sPrometheusCRDAlertingScore {
     pub prometheus_rules: Vec<RuleGroup>,
 }
 
-impl<T: Topology + K8sclient + PrometheusApplicationMonitoring<CRDPrometheus>> Score<T>
+impl<T: Topology + K8sclient + PrometheusMonitoring<CRDPrometheus>> Score<T>
     for K8sPrometheusCRDAlertingScore
 {
     fn create_interpret(&self) -> Box<dyn crate::interpret::Interpret<T>> {
@@ -75,7 +76,7 @@ pub struct K8sPrometheusCRDAlertingInterpret {
 }
 
 #[async_trait]
-impl<T: Topology + K8sclient + PrometheusApplicationMonitoring<CRDPrometheus>> Interpret<T>
+impl<T: Topology + K8sclient + PrometheusMonitoring<CRDPrometheus>> Interpret<T>
     for K8sPrometheusCRDAlertingInterpret
 {
     async fn execute(
@@ -466,10 +467,13 @@ impl K8sPrometheusCRDAlertingInterpret {
             match_labels: label.clone(),
             match_expressions: vec![],
         };
-        let mut json_data = BTreeMap::new();
-        json_data.insert("timeInterval".to_string(), "5s".to_string());
         let namespace = self.sender.namespace.clone();
-
+        let json_data = GrafanaDatasourceJsonData {
+            time_interval: Some("5s".to_string()),
+            http_header_name1: None,
+            tls_skip_verify: Some(true),
+            oauth_pass_thru: Some(true),
+        };
         let json = build_default_dashboard(&namespace);
 
         let graf_data_source = GrafanaDatasource {
@@ -495,7 +499,11 @@ impl K8sPrometheusCRDAlertingInterpret {
                         "http://prometheus-operated.{}.svc.cluster.local:9090",
                         self.sender.namespace.clone()
                     ),
+                    secure_json_data: None,
+                    is_default: None,
+                    editable: None,
                 },
+                values_from: None,
             },
         };
 
@@ -516,7 +524,9 @@ impl K8sPrometheusCRDAlertingInterpret {
             spec: GrafanaDashboardSpec {
                 resync_period: Some("30s".to_string()),
                 instance_selector: labels.clone(),
-                json,
+                json: Some(json),
+                grafana_com: None,
+                datasources: None,
             },
         };
 

harmony/src/modules/prometheus/prometheus.rs

@@ -9,11 +9,17 @@ use crate::{
 };
 
 #[async_trait]
-pub trait PrometheusApplicationMonitoring<S: AlertSender> {
+pub trait PrometheusMonitoring<S: AlertSender> {
     async fn install_prometheus(
         &self,
         sender: &S,
         inventory: &Inventory,
         receivers: Option<Vec<Box<dyn AlertReceiver<S>>>>,
     ) -> Result<PreparationOutcome, PreparationError>;
+
+    async fn ensure_prometheus_operator(
+        &self,
+        sender: &S,
+        inventory: &Inventory,
+    ) -> Result<PreparationOutcome, PreparationError>;
 }

harmony/src/modules/prometheus/rhob_alerting_score.rs

@@ -38,7 +38,7 @@ use crate::{
 };
 use harmony_types::id::Id;
 
-use super::prometheus::PrometheusApplicationMonitoring;
+use super::prometheus::PrometheusMonitoring;
 
 #[derive(Clone, Debug, Serialize)]
 pub struct RHOBAlertingScore {
@@ -48,8 +48,8 @@ pub struct RHOBAlertingScore {
     pub prometheus_rules: Vec<RuleGroup>,
 }
 
-impl<T: Topology + K8sclient + Ingress + PrometheusApplicationMonitoring<RHOBObservability>>
-    Score<T> for RHOBAlertingScore
+impl<T: Topology + K8sclient + Ingress + PrometheusMonitoring<RHOBObservability>> Score<T>
+    for RHOBAlertingScore
 {
     fn create_interpret(&self) -> Box<dyn crate::interpret::Interpret<T>> {
         Box::new(RHOBAlertingInterpret {
@@ -74,8 +74,8 @@ pub struct RHOBAlertingInterpret {
 }
 
 #[async_trait]
-impl<T: Topology + K8sclient + Ingress + PrometheusApplicationMonitoring<RHOBObservability>>
-    Interpret<T> for RHOBAlertingInterpret
+impl<T: Topology + K8sclient + Ingress + PrometheusMonitoring<RHOBObservability>> Interpret<T>
+    for RHOBAlertingInterpret
 {
     async fn execute(
         &self,