Rename health metrics and add resource metrics

api-daemon/pvcapid/flaskapi.py

@@ -640,14 +640,15 @@ class API_Metrics(Resource):
           400:
             description: Bad request
         """
-        cluster_output, cluster_retcode = api_helper.cluster_metrics()
+        health_output, health_retcode = api_helper.cluster_health_metrics()
+        resource_output, resource_retcode = api_helper.cluster_resource_metrics()
         ceph_output, ceph_retcode = api_helper.ceph_metrics()
 
-        if cluster_retcode != 200 or ceph_retcode != 200:
+        if health_retcode != 200 or resource_retcode != 200 or ceph_retcode != 200:
             output = "Error: Failed to obtain data"
             retcode = 400
         else:
-            output = cluster_output + ceph_output
+            output = health_output + resource_output + ceph_output
             retcode = 200
 
         response = flask.make_response(output, retcode)
@@ -658,11 +659,11 @@ class API_Metrics(Resource):
 api.add_resource(API_Metrics, "/metrics")
 
 
-# /metrics/pvc
+# /metrics/health
-class API_Metrics_PVC(Resource):
+class API_Metrics_Health(Resource):
     def get(self):
         """
-        Return the current PVC cluster status in Prometheus-compatible metrics format
+        Return the current PVC cluster health status in Prometheus-compatible metrics format
 
         Endpoint is unauthenticated to allow metrics exfiltration without having to deal
         with the Prometheus compatibility later.
@@ -675,13 +676,13 @@ class API_Metrics_PVC(Resource):
           400:
             description: Bad request
         """
-        cluster_output, cluster_retcode = api_helper.cluster_metrics()
+        health_output, health_retcode = api_helper.cluster_health_metrics()
 
-        if cluster_retcode != 200:
+        if health_retcode != 200:
             output = "Error: Failed to obtain data"
             retcode = 400
         else:
-            output = cluster_output
+            output = health_output
             retcode = 200
 
         response = flask.make_response(output, retcode)
@@ -689,7 +690,41 @@ class API_Metrics_PVC(Resource):
         return response
 
 
-api.add_resource(API_Metrics_PVC, "/metrics/pvc")
+api.add_resource(API_Metrics_Health, "/metrics/health")
+
+
+# /metrics/resource
+class API_Metrics_Resource(Resource):
+    def get(self):
+        """
+        Return the current PVC cluster resource utilizations in Prometheus-compatible metrics format
+
+        Endpoint is unauthenticated to allow metrics exfiltration without having to deal
+        with the Prometheus compatibility later.
+        ---
+        tags:
+          - root
+        responses:
+          200:
+            description: OK
+          400:
+            description: Bad request
+        """
+        resource_output, resource_retcode = api_helper.cluster_resource_metrics()
+
+        if resource_retcode != 200:
+            output = "Error: Failed to obtain data"
+            retcode = 400
+        else:
+            output = resource_output
+            retcode = 200
+
+        response = flask.make_response(output, retcode)
+        response.mimetype = "text/plain"
+        return response
+
+
+api.add_resource(API_Metrics_Resource, "/metrics/resource")
 
 
 # /metrics/ceph

api-daemon/pvcapid/helper.py

@@ -126,12 +126,27 @@ def cluster_maintenance(zkhandler, maint_state="false"):
 #
 @pvc_common.Profiler(config)
 @ZKConnection(config)
-def cluster_metrics(zkhandler):
+def cluster_health_metrics(zkhandler):
     """
-    Format status data from cluster_status into Prometheus-compatible metrics
+    Get cluster-wide Prometheus metrics for health
     """
 
-    retflag, retdata = pvc_cluster.get_metrics(zkhandler)
+    retflag, retdata = pvc_cluster.get_health_metrics(zkhandler)
+    if retflag:
+        retcode = 200
+    else:
+        retcode = 400
+    return retdata, retcode
+
+
+@pvc_common.Profiler(config)
+@ZKConnection(config)
+def cluster_resource_metrics(zkhandler):
+    """
+    Get cluster-wide Prometheus metrics for resource utilization
+    """
+
+    retflag, retdata = pvc_cluster.get_resource_metrics(zkhandler)
     if retflag:
         retcode = 200
     else:

node-daemon/monitoring/prometheus/grafana-pvc-health-dashboard.json

@@ -1510,7 +1510,7 @@
               {
                 "id": "color",
                 "value": {
-                  "fixedColor": "light-red",
+                  "fixedColor": "super-light-red",
                   "mode": "fixed"
                 }
               }
@@ -1525,7 +1525,7 @@
               {
                 "id": "color",
                 "value": {
-                  "fixedColor": "semi-dark-red",
+                  "fixedColor": "red",
                   "mode": "fixed"
                 }
               }
@@ -1540,7 +1540,7 @@
               {
                 "id": "color",
                 "value": {
-                  "fixedColor": "semi-dark-blue",
+                  "fixedColor": "dark-red",
                   "mode": "fixed"
                 }
               }
@@ -2556,7 +2556,8 @@
   "refresh": "5s",
   "schemaVersion": 38,
   "tags": [
-    "pvc"
+    "pvc",
+    "health"
   ],
   "templating": {
     "list": [
@@ -2592,8 +2593,8 @@
   },
   "timepicker": {},
   "timezone": "",
-  "title": "PVC Cluster",
+  "title": "PVC Cluster Health",
   "uid": "fbddd9f9-aadb-4c97-8aea-57c29e5de234",
-  "version": 57,
+  "version": 1,
   "weekStart": ""
 }