fix(observability): bind managed Prometheus datasource for AIO Grafana dashboard

src/000-cloud/020-observability/scripts/import-grafana-dashboards.sh

@@ -8,6 +8,9 @@ set -euo pipefail
 # Environment variables (passed from Terraform)
 GRAFANA_NAME="${GRAFANA_NAME:-}"
 RESOURCE_GROUP_NAME="${RESOURCE_GROUP_NAME:-}"
+# Name of the Azure Monitor Workspace whose auto-provisioned managed Prometheus
+# datasource backs the AIO sample dashboard panels.
+MONITOR_WORKSPACE_NAME="${MONITOR_WORKSPACE_NAME:-}"
 
 if [[ -z "$GRAFANA_NAME" || -z "$RESOURCE_GROUP_NAME" ]]; then
   echo "Error: GRAFANA_NAME and RESOURCE_GROUP_NAME environment variables must be set"
@@ -29,10 +32,10 @@ retry() {
       return 0
     fi
     if ((attempt >= max_attempts)); then
-      echo "Failed after ${max_attempts} attempts"
+      echo "Failed after ${max_attempts} attempts" >&2
       return 1
     fi
-    echo "Attempt ${attempt}/${max_attempts} failed, retrying in ${delay}s..."
+    echo "Attempt ${attempt}/${max_attempts} failed, retrying in ${delay}s..." >&2
     sleep "$delay"
     ((attempt++))
   done
@@ -53,10 +56,52 @@ done
 
 # Import dashboard from GitHub
 echo "Importing AIO sample dashboard from GitHub..."
-retry az grafana dashboard import \
-  -g "$RESOURCE_GROUP_NAME" \
-  -n "$GRAFANA_NAME" \
-  --overwrite \
-  --definition "https://raw.githubusercontent.com/Azure/azure-iot-operations/refs/heads/main/samples/grafana-dashboard/aio.sample.json"
+
+AIO_DASHBOARD_URL="https://raw.githubusercontent.com/Azure/azure-iot-operations/refs/heads/main/samples/grafana-dashboard/aio.sample.json"
+
+# The AIO sample dashboard is a Grafana export: every panel references its
+# datasource through the "${DS_MANAGED_PROMETHEUS_INSTANCE}" __inputs placeholder.
+# az grafana dashboard import only auto-binds that placeholder when a matching
+# Prometheus datasource is Grafana's default. In Azure Managed Grafana the
+# managed Prometheus datasource is not the default, so the placeholder would be
+# imported literally and panels fail with
+# "Datasource ${DS_MANAGED_PROMETHEUS_INSTANCE} was not found".
+# Resolve the managed Prometheus datasource uid and bind it before importing.
+resolve_prometheus_datasource_uid() {
+  az grafana data-source list -g "$RESOURCE_GROUP_NAME" -n "$GRAFANA_NAME" -o json \
+    | jq -r --arg name "$MONITOR_WORKSPACE_NAME" '
+      (map(select(.name == $name)) | .[0].uid)
+      // (map(select(.type == "prometheus")) | .[0].uid)
+      // empty
+    '
+}
+
+DS_UID="$(retry resolve_prometheus_datasource_uid)"
+
+if [[ -z "$DS_UID" ]]; then
+  echo "Warning: could not resolve a managed Prometheus datasource; importing AIO dashboard as-is"
+  retry az grafana dashboard import \
+    -g "$RESOURCE_GROUP_NAME" \
+    -n "$GRAFANA_NAME" \
+    --overwrite \
+    --definition "$AIO_DASHBOARD_URL"
+else
+  echo "Binding AIO dashboard datasource to Prometheus uid: ${DS_UID}"
+  TMP_DIR="$(mktemp -d)"
+  trap 'rm -rf "$TMP_DIR"' EXIT
+  AIO_DASHBOARD_FILE="${TMP_DIR}/aio.sample.json"
+
+  curl -fsSL "$AIO_DASHBOARD_URL" \
+    | jq --arg uid "$DS_UID" '
+      walk(if type == "object" and .uid == "${DS_MANAGED_PROMETHEUS_INSTANCE}" then .uid = $uid else . end)
+      | del(.__inputs)
+    ' >"$AIO_DASHBOARD_FILE"
+
+  retry az grafana dashboard import \
+    -g "$RESOURCE_GROUP_NAME" \
+    -n "$GRAFANA_NAME" \
+    --overwrite \
+    --definition "$AIO_DASHBOARD_FILE"
+fi
 
 echo "Dashboard import completed successfully"

src/000-cloud/020-observability/terraform/main.tf

@@ -130,8 +130,9 @@ resource "terraform_data" "apply_scripts" {
     command = "bash ${path.module}/../scripts/import-grafana-dashboards.sh"
 
     environment = {
-      GRAFANA_NAME        = azurerm_dashboard_grafana.monitor.name
-      RESOURCE_GROUP_NAME = var.azmon_resource_group.name
+      GRAFANA_NAME           = azurerm_dashboard_grafana.monitor.name
+      RESOURCE_GROUP_NAME    = var.azmon_resource_group.name
+      MONITOR_WORKSPACE_NAME = azurerm_monitor_workspace.monitor.name
     }
   }
 }