From 5d4e11ff437fffd3cb91c893b829bb3c3738beb0 Mon Sep 17 00:00:00 2001
From: guptaNswati <swatig@nvidia.com>
Date: Wed, 11 Mar 2020 18:28:40 -0700
Subject: [PATCH] Configure Prometheus to access DCGM metrics

---
 assets/state-monitoring/0400_prom_role.yaml      | 16 ++++++++++++++++
 .../state-monitoring/0500_prom_rolebinding.yaml  | 13 +++++++++++++
 .../{0400_service.yaml => 0600_service.yaml}     |  0
 .../state-monitoring/0700_service_monitor.yaml   | 15 +++++++++++++++
 .../{0500_configmap.yaml => 0800_configmap.yaml} |  0
 .../{0600_daemonset.yaml => 0900_daemonset.yaml} |  0
 .../gpu-operator/templates/0010_namespace.yaml   |  2 ++
 7 files changed, 46 insertions(+)
 create mode 100644 assets/state-monitoring/0400_prom_role.yaml
 create mode 100644 assets/state-monitoring/0500_prom_rolebinding.yaml
 rename assets/state-monitoring/{0400_service.yaml => 0600_service.yaml} (100%)
 create mode 100644 assets/state-monitoring/0700_service_monitor.yaml
 rename assets/state-monitoring/{0500_configmap.yaml => 0800_configmap.yaml} (100%)
 rename assets/state-monitoring/{0600_daemonset.yaml => 0900_daemonset.yaml} (100%)

diff --git a/assets/state-monitoring/0400_prom_role.yaml b/assets/state-monitoring/0400_prom_role.yaml
new file mode 100644
index 000000000..5988aadcf
--- /dev/null
+++ b/assets/state-monitoring/0400_prom_role.yaml
@@ -0,0 +1,16 @@
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: prometheus-k8s
+  namespace: gpu-operator-resources
+rules:
+- apiGroups:
+  - ""
+  resources:
+  - services
+  - endpoints
+  - pods
+  verbs:
+  - get
+  - list
+  - watch
diff --git a/assets/state-monitoring/0500_prom_rolebinding.yaml b/assets/state-monitoring/0500_prom_rolebinding.yaml
new file mode 100644
index 000000000..7b2467780
--- /dev/null
+++ b/assets/state-monitoring/0500_prom_rolebinding.yaml
@@ -0,0 +1,13 @@
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  name: prometheus-k8s
+  namespace: gpu-operator-resources
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: prometheus-k8s
+subjects:
+- kind: ServiceAccount
+  name: prometheus-k8s
+  namespace: openshift-monitoring
diff --git a/assets/state-monitoring/0400_service.yaml b/assets/state-monitoring/0600_service.yaml
similarity index 100%
rename from assets/state-monitoring/0400_service.yaml
rename to assets/state-monitoring/0600_service.yaml
diff --git a/assets/state-monitoring/0700_service_monitor.yaml b/assets/state-monitoring/0700_service_monitor.yaml
new file mode 100644
index 000000000..13a1858ac
--- /dev/null
+++ b/assets/state-monitoring/0700_service_monitor.yaml
@@ -0,0 +1,15 @@
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+  name: nvidia-dcgm-exporter
+  namespace: gpu-operator-resources
+spec:
+  endpoints:
+  - port: gpu-metrics
+  jobLabel: app
+  namespaceSelector:
+    matchNames:
+    - gpu-operator-resources
+  selector:
+    matchLabels:
+     app: nvidia-dcgm-exporter
diff --git a/assets/state-monitoring/0500_configmap.yaml b/assets/state-monitoring/0800_configmap.yaml
similarity index 100%
rename from assets/state-monitoring/0500_configmap.yaml
rename to assets/state-monitoring/0800_configmap.yaml
diff --git a/assets/state-monitoring/0600_daemonset.yaml b/assets/state-monitoring/0900_daemonset.yaml
similarity index 100%
rename from assets/state-monitoring/0600_daemonset.yaml
rename to assets/state-monitoring/0900_daemonset.yaml
diff --git a/deployments/gpu-operator/templates/0010_namespace.yaml b/deployments/gpu-operator/templates/0010_namespace.yaml
index 8e65a7d28..ac2a14be9 100644
--- a/deployments/gpu-operator/templates/0010_namespace.yaml
+++ b/deployments/gpu-operator/templates/0010_namespace.yaml
@@ -2,3 +2,5 @@ apiVersion: v1
 kind: Namespace
 metadata:
     name: gpu-operator-resources
+    labels:
+      openshift.io/cluster-monitoring: "true"