Add kubernetes persistentvolume metrics

kubernetes_state/datadog_checks/kubernetes_state/kubernetes_state.py

@@ -73,6 +73,7 @@ def __init__(self, name, init_config, agentConfig, instances=None):
             'kube_pod_container_status_terminated': 'container.terminated',
             'kube_pod_container_status_waiting': 'container.waiting',
             'kube_persistentvolumeclaim_status_phase': 'persistentvolumeclaim.status',
+            'kube_persistentvolumeclaim_resource_requests_storage_bytes': 'persistentvolumeclaim.request_storage',
             'kube_pod_container_resource_limits_cpu_cores': 'container.cpu_limit',
             'kube_pod_container_resource_limits_memory_bytes': 'container.memory_limit',
             'kube_pod_container_resource_requests_cpu_cores': 'container.cpu_requested',
@@ -161,6 +162,14 @@ def __init__(self, name, init_config, agentConfig, instances=None):
             'kube_pod_info': {
                 'label_to_match': 'pod',
                 'labels_to_get': ['node']
+            },
+            'kube_persistentvolume_info': {
+                'label_to_match': 'persistentvolume',
+                'labels_to_get': ['storageclass']
+            },
+            'kube_persistentvolumeclaim_info': {
+                'label_to_match': 'persistentvolumeclaim',
+                'labels_to_get': ['storageclass']
             }
         }
 
@@ -581,3 +590,18 @@ def kube_limitrange(self, message, **kwargs):
                 self.gauge(metric_base_name.format(resource, constraint), val, tags)
         else:
             self.log.error("Metric type %s unsupported for metric %s" % (message.type, message.name))
+
+    def kube_persistentvolume_status_phase(self, message, **kwargs):
+        """ The persistent volumes by phase. """
+        metric_name = self.NAMESPACE + '.persistentvolumes.by_phase'
+        by_phase_counter = Counter()
+
+        for metric in message.metric:
+            tags = [
+                self._label_to_tag("storageclass", metric.label),
+                self._label_to_tag("phase", metric.label)
+            ] + self.custom_tags
+            by_phase_counter[tuple(sorted(tags))] += metric.gauge.value
+
+        for tags, count in by_phase_counter.iteritems():
+            self.gauge(metric_name, count, tags=list(tags))

kubernetes_state/metadata.csv

@@ -40,7 +40,9 @@ kubernetes_state.node.memory_capacity,gauge,,byte,,The total memory resources of
 kubernetes_state.node.pods_capacity,gauge,,,,The total pod resources of the node,0,kubernetes,k8s_state.node.pods_capacity
 kubernetes_state.node.gpu.cards_allocatable,gauge,,,,The GPU resources of a node that are available for scheduling,0,kubernetes,k8s_state.node.gpu.cards_allocatable
 kubernetes_state.node.gpu.cards_capacity,gauge,,,,The total GPU resources of the node,0,kubernetes,k8s_state.node.gpu.cards_capacity
-kubernetes_state.persistentvolumeclaim.status,gauge,,,,The phase the persistent volume claim is currently in,-1,kubernetes,k8s_state.persistentvolumeclaim.status
+kubernetes_state.persistentvolumeclaim.status,gauge,,,,The phase the persistent volume claim is currently in,-1,kubernetes,k8s_state.pvc.status
+kubernetes_state.persistentvolumeclaim.request_storage,gauge,,byte,,Storage space request for a given pvc,0,kubernetes,k8s_state.pvc.request
+kubernetes_state.persistentvolume.by_phase,gauge,,,,Number of persistent volumes to sum by phase and storageclass,0,kubernetes,k8s_state.pv.by_phase
 kubernetes_state.node.cpu_allocatable,gauge,,cpu,,The CPU resources of a node that are available for scheduling,0,kubernetes,k8s_state.node.cpu_allocatable
 kubernetes_state.node.memory_allocatable,gauge,,byte,,The memory resources of a node that are available for scheduling,0,kubernetes,k8s_state.node.memory_allocatable
 kubernetes_state.node.pods_allocatable,gauge,,,,The pod resources of a node that are available for scheduling,0,kubernetes,k8s_state.node.pods_allocatable

kubernetes_state/tests/fixtures/prometheus.txt

@@ -258,6 +258,21 @@ kube_node_status_condition{condition="Ready",node="minikube",status="unknown"} 0
 # HELP kube_node_status_network_unavailable Whether the network is correctly configured for the node.
 # TYPE kube_node_status_network_unavailable gauge
 kube_node_status_network_unavailable{node="127.0.0.1",condition="false"} 1
+# TYPE kube_persistentvolume_status_phase gauge
+kube_persistentvolume_status_phase{persistentvolume="local-pv-103fef5d",phase="Available"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-103fef5d",phase="Bound"} 1
+kube_persistentvolume_status_phase{persistentvolume="local-pv-103fef5d",phase="Failed"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-103fef5d",phase="Pending"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-103fef5d",phase="Released"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-15d0a3a3",phase="Available"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-15d0a3a3",phase="Bound"} 1
+kube_persistentvolume_status_phase{persistentvolume="local-pv-15d0a3a3",phase="Failed"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-15d0a3a3",phase="Pending"} 0
+kube_persistentvolume_status_phase{persistentvolume="local-pv-15d0a3a3",phase="Released"} 0
+# HELP kube_persistentvolume_info Information about persistentvolume.
+# TYPE kube_persistentvolume_info gauge
+kube_persistentvolume_info{persistentvolume="local-pv-103fef5d",storageclass="local-data"} 1
+kube_persistentvolume_info{persistentvolume="local-pv-15d0a3a3",storageclass="local-data"} 1
 # HELP kube_persistentvolumeclaim_info Information about persistent volume claim.
 # TYPE kube_persistentvolumeclaim_info gauge
 kube_persistentvolumeclaim_info{namespace="default",persistentvolumeclaim="task-pv-claim",storageclass="manual"} 1

kubernetes_state/tests/test_kubernetes_state.py

@@ -68,6 +68,7 @@
     NAMESPACE + '.replicaset.replicas_desired',
     # persistentvolume claim
     NAMESPACE + '.persistentvolumeclaim.status',
+    NAMESPACE + '.persistentvolumeclaim.request_storage',
     # statefulset
     NAMESPACE + '.statefulset.replicas',
     NAMESPACE + '.statefulset.replicas_current',
@@ -94,6 +95,9 @@
         'reason:CrashLoopBackOff',
         'reason:ErrImagePull',
         'reason:ImagePullBackoff'
+    ],
+    NAMESPACE + '.persistentvolumeclaim.request_storage': [
+        'storageclass:manual'
     ]
 }
 
@@ -237,6 +241,18 @@ def test_update_kube_state_metrics(aggregator, instance, check):
     aggregator.assert_metric(NAMESPACE + '.pod.status_phase',
                              tags=['namespace:default', 'phase:Unknown', 'optional:tag1'], value=1)
 
+    # Persistentvolume counts
+    aggregator.assert_metric(NAMESPACE + '.persistentvolumes.by_phase',
+                             tags=['storageclass:local-data', 'phase:Available', 'optional:tag1'], value=0)
+    aggregator.assert_metric(NAMESPACE + '.persistentvolumes.by_phase',
+                             tags=['storageclass:local-data', 'phase:Bound', 'optional:tag1'], value=2)
+    aggregator.assert_metric(NAMESPACE + '.persistentvolumes.by_phase',
+                             tags=['storageclass:local-data', 'phase:Failed', 'optional:tag1'], value=0)
+    aggregator.assert_metric(NAMESPACE + '.persistentvolumes.by_phase',
+                             tags=['storageclass:local-data', 'phase:Pending', 'optional:tag1'], value=0)
+    aggregator.assert_metric(NAMESPACE + '.persistentvolumes.by_phase',
+                             tags=['storageclass:local-data', 'phase:Released', 'optional:tag1'], value=0)
+
     for metric in METRICS:
         aggregator.assert_metric(metric, hostname=HOSTNAMES.get(metric, None))
         for tag in TAGS.get(metric, []):