[otel-integration] remove kube-state-metrics (#326)

otel-integration/CHANGELOG.md

@@ -2,6 +2,9 @@
 
 ## OpenTelemtry-Integration
 
+### v0.0.33 / 2023-11-08
+* [FIX] Remove Kube-State-Metrics, as K8s Cluster Receiver provides all the needed metrics.
+
 ### v0.0.32 / 2023-11-03
 * [FIX] Ensure correct order of processors for k8s deployment attributes.
 

otel-integration/k8s-helm/Chart.yaml

@@ -1,7 +1,7 @@
 apiVersion: v2
 name: otel-integration
 description: OpenTelemetry Integration
-version: 0.0.32
+version: 0.0.33
 keywords:
   - OpenTelemetry Collector
   - OpenTelemetry Agent
@@ -24,10 +24,6 @@ dependencies:
     version: "0.76.3"
     repository: https://cgx.jfrog.io/artifactory/coralogix-charts-virtual
     condition: opentelemetry-cluster-collector.enabled
-  - name: kube-state-metrics
-    version: "5.8.1"
-    repository: https://prometheus-community.github.io/helm-charts
-    condition: global.extensions.kubernetesDashboard.enabled
 sources:
   - https://github.com/coralogix/opentelemetry-helm-charts/tree/main/charts/opentelemetry-collector
 maintainers:

otel-integration/k8s-helm/values.yaml

@@ -329,6 +329,15 @@ opentelemetry-cluster-collector:
       k8s_cluster:
         collection_interval: 10s
         allocatable_types_to_report: [cpu, memory]
+        resource_attributes:
+          k8s.kubelet.version:
+            enabled: true
+          k8s.pod.qos_class:
+            enabled: true
+        metrics:
+          k8s.pod.status_reason:
+            enabled: true
+
       prometheus:
         config:
           scrape_configs:
@@ -367,22 +376,51 @@ opentelemetry-cluster-collector:
               - keep_keys(body["object"], ["type", "eventTime", "reason", "regarding", "note", "metadata", "deprecatedFirstTimestamp", "deprecatedLastTimestamp"])
               - keep_keys(body["object"]["metadata"], ["creationTimestamp"])
               - keep_keys(body["object"]["regarding"], ["kind", "name", "namespace"])
-      metricstransform/kube-extra-metrics:
+      metricstransform/k8s-dashboard:
         transforms:
-          # Replace node name for kube node info with the name of the target node.
-        - include: kube_node_info
+        - include: k8s.pod.phase
+          match_type: strict
+          action: insert
+          new_name: kube_pod_status_qos_class
+        - include: k8s.pod.status_reason
+          match_type: strict
+          action: insert
+          new_name: kube_pod_status_reason
+        - include: k8s.node.allocatable_cpu
           match_type: strict
-          action: update
-          operations:
-            - action: update_label
-              label: node
-              new_label: k8s.node.name
+          action: insert
+          new_name: kube_node_info
       transform/k8s-dashboard:
         error_mode: ignore
         metric_statements:
           - context: metric
             statements:
+              # k8s.pod.phase has changed metric units to follow OTEL semantic conventions
+              # K8s Dashboard uses k8s_pod_phase_1 in their queries.
               - set(unit, "1") where name == "k8s.pod.phase"
+              - set(unit, "") where name == "kube_node_info"
+          - context: datapoint
+            statements:
+              # Transforming k8s.pod.phase to kube_pod_status_qos_class format.
+              - set(value_int, 1) where metric.name == "kube_pod_status_qos_class"
+              - set(attributes["qos_class"], resource.attributes["k8s.pod.qos_class"]) where metric.name == "kube_pod_status_qos_class"
+              - set(attributes["pod"], resource.attributes["k8s.pod.name"]) where metric.name == "kube_pod_status_reason"
+              # Transforming k8s.pod.status_reason to kube-state-metrics format
+              - set(attributes["reason"], "Evicted") where metric.name == "kube_pod_status_reason" and value_int == 1
+              - set(attributes["reason"], "NodeAffinity") where metric.name == "kube_pod_status_reason" and value_int == 2
+              - set(attributes["reason"], "NodeLost") where metric.name == "kube_pod_status_reason" and value_int == 3
+              - set(attributes["reason"], "Shutdown") where metric.name == "kube_pod_status_reason" and value_int == 4
+              - set(attributes["reason"], "UnexpectedAdmissionError") where metric.name == "kube_pod_status_reason" and value_int == 5
+              - set(value_int, 0) where metric.name == "kube_pod_status_reason" and value_int == 6
+              - set(value_int, 1) where metric.name == "kube_pod_status_reason" and value_int != 0
+              # Transforming k8s.node.status_reason to kube-state-metrics format
+              - set(value_int, 1) where metric.name == "kube_node_info"
+              - set(attributes["kubelet_version"], resource.attributes["k8s.kubelet.version"]) where metric.name == "kube_node_info"
+          - context: resource
+            statements:
+              - delete_key(attributes, "k8s.pod.qos_class")
+              - delete_key(attributes, "k8s.kubelet.version")
+
       resourcedetection/env:
         detectors: ["system", "env"]
         timeout: 2s
@@ -437,7 +475,7 @@ opentelemetry-cluster-collector:
             - coralogix
           processors:
             - k8sattributes
-            - metricstransform/kube-extra-metrics
+            - metricstransform/k8s-dashboard
             - transform/k8s-dashboard
             - resourcedetection/env
             - resourcedetection/region
@@ -485,15 +523,3 @@ opentelemetry-cluster-collector:
 
 opentelemetry-agent-windows:
   enabled: false
-
-kube-state-metrics:
-  fullnameOverride: coralogix-opentelemetry-kube-state-metrics
-  prometheusScrape: false
-  collectors:
-    - pods
-    - nodes
-  metricAllowlist:
-    - kube_node_info
-    - kube_pod_status_reason
-    - kube_pod_status_phase
-    - kube_pod_status_qos_class