charts/meta-monitoring/Chart.lock

@@ -1,18 +1,18 @@
 dependencies:
 - name: loki
   repository: https://grafana.github.io/helm-charts
-  version: 6.29.0
+  version: 6.6.2
 - name: alloy
   repository: https://grafana.github.io/helm-charts
-  version: 0.12.5
+  version: 0.3.2
 - name: mimir-distributed
   repository: https://grafana.github.io/helm-charts
-  version: 5.6.0
+  version: 5.3.0
 - name: tempo-distributed
   repository: https://grafana.github.io/helm-charts
-  version: 1.33.0
+  version: 1.10.0
 - name: minio
   repository: https://charts.min.io
-  version: 5.4.0
+  version: 5.2.0
-digest: sha256:5225a03d9003384639f5d43b1971126371269347f16f221b7aed377ab85d71be
+digest: sha256:f62b04092f6e7e89e6a427b586e4c0cbfcd54ddbd411846770ee684cf86f822a
-generated: "2025-03-27T07:03:11.17404081Z"
+generated: "2024-06-03T07:02:47.330637227Z"

charts/meta-monitoring/Chart.yaml

@@ -13,7 +13,7 @@ type: application
 # This is the chart version. This version number should be incremented each time you make changes
 # to the chart and its templates, including the app version.
 # Versions are expected to follow Semantic Versioning (https://semver.org/)
-version: 1.3.0
+version: 1.0.0
 # This is the version number of the application being deployed. This version number should be
 # incremented each time you make changes to the application. Versions are not expected to
 # follow Semantic Versioning. They should reflect the version the application is using.
@@ -22,20 +22,20 @@ appVersion: "0.0.1"
 dependencies:
 - name: loki
   repository: https://grafana.github.io/helm-charts
-  version: 6.29.0
+  version: 6.6.2
   condition: local.logs.enabled
 - name: alloy
   repository: https://grafana.github.io/helm-charts
-  version: 0.12.5
+  version: 0.3.2
 - name: mimir-distributed
   repository: https://grafana.github.io/helm-charts
-  version: 5.6.0
+  version: 5.3.0
   condition: local.metrics.enabled
 - name: tempo-distributed
   repository: https://grafana.github.io/helm-charts
-  version: 1.33.0
+  version: 1.10.0
   condition: local.traces.enabled
 - name: minio
   repository: https://charts.min.io
-  version: 5.4.0
+  version: 5.2.0
   condition: local.minio.enabled

charts/meta-monitoring/templates/agent/config.yaml

@@ -155,7 +155,133 @@ data:
 
       forward_to = [ {{ include "agent.prometheus_write_targets" . }} ]
     }
+    {{- if .Values.kubeStateMetrics.enabled }}
 
+    prometheus.scrape "kubeStateMetrics" {
+      clustering {
+        enabled = true
+      }
+      targets    = [ { "__address__" = "{{ .Values.kubeStateMetrics.endpoint }}" } ]
+      forward_to = [ prometheus.relabel.filter.receiver ]
+    }
+    {{- end }}
+
+    // cAdvisor and Kubelet metrics
+    // Based on https://github.com/Chewie/loutretelecom-manifests/blob/main/manifests/addons/monitoring/config.river
+    discovery.kubernetes "all_nodes" {
+      role = "node"
+      namespaces {
+        own_namespace = true
+        names = [ {{ include "agent.namespaces" . }} ]
+      }
+    }
+
+    discovery.relabel "all_nodes" {
+      targets = discovery.kubernetes.all_nodes.targets
+      rule {
+        source_labels = ["__meta_kubernetes_node_name"]
+        target_label = "node"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_namespace"]
+        target_label  = "namespace"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_pod_name"]
+        target_label  = "pod"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_namespace", "__meta_kubernetes_pod_label_app_kubernetes_io_name", "__meta_kubernetes_pod_label_app_kubernetes_io_component"]
+        separator = "/"
+        regex = "(.*)/(.*)/(.*)"
+        replacement = "${1}/${2}-${3}"
+        target_label = "job"
+      }
+      rule {
+        target_label = "cluster"
+        replacement = "{{- .Values.clusterLabelValue -}}"
+      }
+    }
+
+    prometheus.scrape "cadvisor" {
+      clustering {
+        enabled = true
+      }
+      targets    = discovery.relabel.all_nodes.output
+      forward_to = [ prometheus.relabel.filter.receiver ]
+
+      metrics_path = "/metrics/cadvisor"
+      scheme = "https"
+
+      bearer_token_file = "/var/run/secrets/kubernetes.io/serviceaccount/token"
+      tls_config {
+        ca_file = "/var/run/secrets/kubernetes.io/serviceaccount/ca.crt"
+      }
+    }
+
+    prometheus.scrape "kubelet" {
+      clustering {
+        enabled = true
+      }
+      targets    = discovery.relabel.all_nodes.output
+      forward_to = [ prometheus.relabel.filter.receiver ]
+
+      metrics_path = "/metrics"
+      scheme = "https"
+
+      bearer_token_file = "/var/run/secrets/kubernetes.io/serviceaccount/token"
+      tls_config {
+        ca_file = "/var/run/secrets/kubernetes.io/serviceaccount/ca.crt"
+      }
+    }
+
+    prometheus.exporter.unix "promexporter" {}
+
+    prometheus.scrape "node_exporter" {
+      clustering {
+        enabled = true
+      }
+      targets = prometheus.exporter.unix.promexporter.targets
+      forward_to = [prometheus.relabel.node_exporter.receiver]
+
+      job_name = "node-exporter"
+    }
+
+    prometheus.relabel "node_exporter" {
+      forward_to = [ prometheus.relabel.filter.receiver ]
+
+      rule {
+        replacement = env("HOSTNAME")
+        target_label = "nodename"
+      }
+      rule {
+        replacement = "node-exporter"
+        target_label = "job"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_node_name"]
+        target_label = "node"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_namespace"]
+        target_label  = "namespace"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_pod_name"]
+        target_label  = "pod"
+      }
+      rule {
+        source_labels = ["__meta_kubernetes_namespace", "__meta_kubernetes_pod_label_app_kubernetes_io_name", "__meta_kubernetes_pod_label_app_kubernetes_io_component"]
+        separator = "/"
+        regex = "(.*)/(.*)/(.*)"
+        replacement = "${1}/${2}-${3}"
+        target_label = "job"
+      }
+      rule {
+        target_label = "cluster"
+        replacement = "{{- .Values.clusterLabelValue -}}"
+      }
+    }
     {{- end }}
 
     {{- if or .Values.local.traces.enabled .Values.cloud.traces.enabled }}

charts/meta-monitoring/templates/validate.yaml

@@ -34,6 +34,10 @@
 {{- end -}}
 {{- end -}}
 
+{{- if empty .Values.namespacesToMonitor -}}
+  {{- fail "No namespaces have been specified in namespacesToMonitor" -}}
+{{- end -}}
+
 {{- if empty .Values.metrics.retain -}}
   {{- fail "All metrics will be collected, please specify some in metrics.retain" -}}
 {{- end -}}

charts/meta-monitoring/values.yaml

@@ -1,7 +1,6 @@
 # Specify the namespaces to monitor here
-# By default the chart will monitor the namespace it is installed in
+namespacesToMonitor:
-# namespacesToMonitor:
+- loki
-# - loki
 # The name of the cluster where this will be installed
 clusterLabelValue: "meta"
 # Set to true to write logs, metrics or traces to Grafana Cloud
@@ -29,7 +28,7 @@ local:
   minio:
     enabled: false # This should be set to true if any of the previous is enabled
 grafana:
-  version: 11.4.3
+  version: 10.4.2
   # Gateway ingress configuration
   ingress:
     # -- Specifies whether an ingress for the gateway should be created
@@ -61,10 +60,10 @@ logs:
   # Adding regexes here will add a stage.replace block for logs. For more information see
   # https://grafana.com/docs/agent/latest/flow/reference/components/loki.process/#stagereplace-block
   piiRegexes: null # This example replaces the word after password with *****
-  # - expression: "password (\\\\S+)"
+# - expression: "password (\\\\S+)"
-  #   source: ""         # Empty uses the log message
+#   source: ""         # Empty uses the log message
-  #   replace: "*****""
+#   replace: "*****""
-  # The lines matching these will be kept in Loki
+# The lines matching these will be kept in Loki
   retain:
   # This shows the queries
   - executing query