use kube_deployment_status_observed_generation to properly count depl…

…oyments
DataDog · Dec 21, 2020 · 24d413e · 24d413e
1 parent fcf960d
commit 24d413e
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 8 deletions.
diff --git a/kubernetes_state/datadog_checks/kubernetes_state/kubernetes_state.py b/kubernetes_state/datadog_checks/kubernetes_state/kubernetes_state.py
@@ -101,9 +101,9 @@ def __init__(self, name, init_config, instances):
                 'allowed_labels': ['namespace', 'owner_name', 'owner_kind'],
             },
             'kube_job_owner': {'metric_name': 'job.count', 'allowed_labels': ['namespace', 'owner_name', 'owner_kind']},
-            'kube_deployment_status_condition': {
+            'kube_deployment_status_observed_generation': {
                 'metric_name': 'deployment.count',
-                'allowed_labels': ['namespace', 'condition', 'status'],
+                'allowed_labels': ['namespace'],
             },
         }
 
@@ -131,7 +131,7 @@ def __init__(self, name, init_config, instances):
             'kube_replicaset_owner': self.count_objects_by_tags,
             'kube_job_owner': self.count_objects_by_tags,
             # to get overall count is to filter by Available
-            'kube_deployment_status_condition': self.count_objects_by_tags,
+            'kube_deployment_status_observed_generation': self.count_objects_by_tags,
         }
 
         # Handling cron jobs succeeded/failed counts
@@ -318,7 +318,7 @@ def _create_kubernetes_state_prometheus_instance(self, instance):
                     # _generation metrics are more metadata than metrics, no real use case for now
                     'kube_daemonset_metadata_generation',
                     'kube_deployment_metadata_generation',
-                    'kube_deployment_status_observed_generation',
+                    'kube_deployment_status_condition',
                     'kube_replicaset_metadata_generation',
                     'kube_replicaset_status_observed_generation',
                     'kube_replicationcontroller_metadata_generation',

diff --git a/kubernetes_state/tests/test_kubernetes_state.py b/kubernetes_state/tests/test_kubernetes_state.py
@@ -27,7 +27,6 @@
     NAMESPACE + '.node.gpu.cards_allocatable',
     NAMESPACE + '.nodes.by_condition',
     # deployments
-    NAMESPACE + '.deployment.count',
     NAMESPACE + '.deployment.replicas',
     NAMESPACE + '.deployment.replicas_available',
     NAMESPACE + '.deployment.replicas_unavailable',
@@ -149,7 +148,6 @@
         'namespace:kube-system',
     ],
     NAMESPACE + '.pod.count': ['uid:b6fb4273-2dd6-4edb-9a23-7642bb121806', 'created_by_kind:daemonset'],
-    NAMESPACE + '.deployment.count': ['condition:progressing', 'condition:available', 'status:true'],
     NAMESPACE + '.replicaset.count': ['owner_kind:deployment', 'owner_name:metrics-server-v0.3.6'],
     NAMESPACE + '.namespace.count': ['phase:active', 'phase:terminating'],
     NAMESPACE + '.job.count': ['owner_kind:cronjob', 'owner_name:a-cronjob'],
@@ -375,6 +373,18 @@ def test_update_kube_state_metrics(aggregator, instance, check):
         value=0,
     )
 
+    # deployment counts
+    aggregator.assert_metric(
+        NAMESPACE + '.deployment.count',
+        tags=['namespace:default', 'optional:tag1'],
+        value=2,
+    )
+    aggregator.assert_metric(
+        NAMESPACE + '.deployment.count',
+        tags=['namespace:kube-system', 'optional:tag1'],
+        value=2,
+    )
+
     for metric in METRICS:
         aggregator.assert_metric(metric, hostname=HOSTNAMES.get(metric, None))
         for tag in TAGS.get(metric, []):
@@ -806,10 +816,10 @@ def test_telemetry(aggregator, instance):
     for _ in range(2):
         check.check(instance)
     aggregator.assert_metric(NAMESPACE + '.telemetry.payload.size', tags=['optional:tag1'], value=93895.0)
-    aggregator.assert_metric(NAMESPACE + '.telemetry.metrics.processed.count', tags=['optional:tag1'], value=998.0)
+    aggregator.assert_metric(NAMESPACE + '.telemetry.metrics.processed.count', tags=['optional:tag1'], value=994.0)
     aggregator.assert_metric(NAMESPACE + '.telemetry.metrics.input.count', tags=['optional:tag1'], value=1326.0)
     aggregator.assert_metric(NAMESPACE + '.telemetry.metrics.blacklist.count', tags=['optional:tag1'], value=24.0)
-    aggregator.assert_metric(NAMESPACE + '.telemetry.metrics.ignored.count', tags=['optional:tag1'], value=328.0)
+    aggregator.assert_metric(NAMESPACE + '.telemetry.metrics.ignored.count', tags=['optional:tag1'], value=332.0)
     aggregator.assert_metric(
         NAMESPACE + '.telemetry.collector.metrics.count',
         tags=['resource_name:pod', 'resource_namespace:default', 'optional:tag1'],