Add extra fields for keda triggers and add authentication

grafana · Dec 30, 2024 · 2484213 · 2484213
1 parent f8e07d0
commit 2484213
Show file tree

Hide file tree

Showing 8 changed files with 114 additions and 10 deletions.
diff --git a/operations/helm/charts/mimir-distributed/templates/distributor/distributor-so.yaml b/operations/helm/charts/mimir-distributed/templates/distributor/distributor-so.yaml
@@ -25,20 +25,32 @@ spec:
   triggers:
   - metadata:
       query: max_over_time(sum(sum by (pod) (rate(container_cpu_usage_seconds_total{container="distributor",namespace="{{ .Release.Namespace }}"}[5m])) and max by (pod) (up{container="distributor",namespace="{{ .Release.Namespace }}"}) > 0)[15m:]) * 1000
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       {{- $cpu_request := dig "requests" "cpu" nil .Values.distributor.resources }}
       threshold: {{ mulf (include "mimir.cpuToMilliCPU" (dict "value" $cpu_request)) (divf .Values.distributor.kedaAutoscaling.targetCPUUtilizationPercentage 100) | floor | int64 | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   - metadata:
       query: max_over_time(sum((sum by (pod) (container_memory_working_set_bytes{container="distributor",namespace="{{ .Release.Namespace }}"}) and max by (pod) (up{container="distributor",namespace="{{ .Release.Namespace }}"}) > 0) or vector(0))[15m:]) + sum(sum by (pod) (max_over_time(kube_pod_container_resource_requests{container="distributor",namespace="{{ .Release.Namespace }}", resource="memory"}[15m])) and max by (pod) (changes(kube_pod_container_status_restarts_total{container="distributor",namespace="{{ .Release.Namespace }}"}[15m]) > 0) and max by (pod) (kube_pod_container_status_last_terminated_reason{container="distributor",namespace="{{ .Release.Namespace }}", reason="OOMKilled"}) or vector(0))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       {{- $mem_request := dig "requests" "memory" nil .Values.distributor.resources }}
       threshold: {{ mulf (include "mimir.siToBytes" (dict "value" $mem_request)) (divf .Values.distributor.kedaAutoscaling.targetMemoryUtilizationPercentage 100) | floor | int64 | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
 {{- end }}
diff --git a/operations/helm/charts/mimir-distributed/templates/keda/keda-trigger-auth.yaml b/operations/helm/charts/mimir-distributed/templates/keda/keda-trigger-auth.yaml
@@ -0,0 +1,12 @@
+{{- if .Values.kedaAutoscaling.authentication.enabled }}
+apiVersion: keda.sh/v1alpha1
+kind: TriggerAuthentication
+metadata:
+  name: keda-triggger-auth
+  namespace: {{ .Release.Namespace | quote }}
+spec:
+  secretTargetRef:
+  {{- with .Values.kedaAutoscaling.authentication.secretTargetRef }}
+    {{- toYaml . | nindent 2 }}
+  {{- end }}
+{{- end }}
diff --git a/operations/helm/charts/mimir-distributed/templates/querier/querier-so.yaml b/operations/helm/charts/mimir-distributed/templates/querier/querier-so.yaml
@@ -28,22 +28,34 @@ spec:
   triggers:
   - metadata:
       query: sum(max_over_time(cortex_query_scheduler_inflight_requests{container="query-scheduler",namespace="{{ .Release.Namespace }}",quantile="0.5"}[1m]))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       threshold: {{ .Values.querier.kedaAutoscaling.querySchedulerInflightRequestsThreshold | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     name: cortex_querier_hpa_default
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   - metadata:
       query: sum(rate(cortex_querier_request_duration_seconds_sum{container="querier",namespace="{{ .Release.Namespace }}"}[1m]))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       threshold: {{ .Values.querier.kedaAutoscaling.querySchedulerInflightRequestsThreshold | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     name: cortex_querier_hpa_default_requests_duration
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   {{- $autoscaling := .Values.querier.kedaAutoscaling -}}
   {{- if .Values.querier.kedaAutoscaling.predictiveScalingEnabled }}
   - metadata:
@@ -53,7 +65,13 @@ spec:
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     name: cortex_querier_hpa_default_predictive
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   {{- end }}
 {{- end }}
diff --git a/operations/helm/charts/mimir-distributed/templates/query-frontend/query-frontend-so.yaml b/operations/helm/charts/mimir-distributed/templates/query-frontend/query-frontend-so.yaml
@@ -25,20 +25,32 @@ spec:
   triggers:
   - metadata:
       query: max_over_time(sum(sum by (pod) (rate(container_cpu_usage_seconds_total{container="query-frontend",namespace="{{ .Release.Namespace }}"}[5m])) and max by (pod) (up{container="query-frontend",namespace="{{ .Release.Namespace }}"}) > 0)[15m:]) * 1000
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       {{- $cpu_request := dig "requests" "cpu" nil .Values.query_frontend.resources }}
       threshold: {{ mulf (include "mimir.cpuToMilliCPU" (dict "value" $cpu_request)) (divf .Values.query_frontend.kedaAutoscaling.targetCPUUtilizationPercentage 100) | floor | int64 | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   - metadata:
       query: max_over_time(sum((sum by (pod) (container_memory_working_set_bytes{container="query-frontend",namespace="{{ .Release.Namespace }}"}) and max by (pod) (up{container="query-frontend",namespace="{{ .Release.Namespace }}"}) > 0) or vector(0))[15m:]) + sum(sum by (pod) (max_over_time(kube_pod_container_resource_requests{container="query-frontend",namespace="{{ .Release.Namespace }}", resource="memory"}[15m])) and max by (pod) (changes(kube_pod_container_status_restarts_total{container="query-frontend",namespace="{{ .Release.Namespace }}"}[15m]) > 0) and max by (pod) (kube_pod_container_status_last_terminated_reason{container="query-frontend",namespace="{{ .Release.Namespace }}", reason="OOMKilled"}) or vector(0))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       {{- $mem_request := dig "requests" "memory" nil .Values.query_frontend.resources }}
       threshold: {{ mulf (include "mimir.siToBytes" (dict "value" $mem_request)) (divf .Values.query_frontend.kedaAutoscaling.targetMemoryUtilizationPercentage 100) | floor | int64 | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
 {{- end }}
diff --git a/operations/helm/charts/mimir-distributed/templates/ruler-querier/ruler-querier-so.yaml b/operations/helm/charts/mimir-distributed/templates/ruler-querier/ruler-querier-so.yaml
@@ -26,21 +26,33 @@ spec:
   triggers:
   - metadata:
       query: sum(max_over_time(cortex_query_scheduler_inflight_requests{container="ruler-query-scheduler",namespace="{{ .Release.Namespace }}",quantile="0.5"}[1m]))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       threshold: {{ .Values.ruler_querier.kedaAutoscaling.querySchedulerInflightRequestsThreshold | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     name: cortex_querier_hpa_default
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   - metadata:
       query: sum(rate(cortex_querier_request_duration_seconds_sum{container="ruler-querier",namespace="{{ .Release.Namespace }}"}[1m]))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       threshold: {{ .Values.ruler_querier.kedaAutoscaling.querySchedulerInflightRequestsThreshold | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     name: cortex_querier_hpa_default_requests_duration
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
 {{- end }}
 {{- end }}
diff --git a/...helm/charts/mimir-distributed/templates/ruler-query-frontend/ruler-query-frontend-so.yaml b/...helm/charts/mimir-distributed/templates/ruler-query-frontend/ruler-query-frontend-so.yaml
@@ -32,7 +32,13 @@ spec:
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   - metadata:
       query: max_over_time(sum((sum by (pod) (container_memory_working_set_bytes{container="ruler-query-frontend",namespace="{{ .Release.Namespace }}"}) and max by (pod) (up{container="ruler-query-frontend",namespace="{{ .Release.Namespace }}"}) > 0) or vector(0))[15m:]) + sum(sum by (pod) (max_over_time(kube_pod_container_resource_requests{container="ruler-query-frontend",namespace="{{ .Release.Namespace }}", resource="memory"}[15m])) and max by (pod) (changes(kube_pod_container_status_restarts_total{container="ruler-query-frontend",namespace="{{ .Release.Namespace }}"}[15m]) > 0) and max by (pod) (kube_pod_container_status_last_terminated_reason{container="ruler-query-frontend",namespace="{{ .Release.Namespace }}", reason="OOMKilled"}) or vector(0))
       serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
@@ -41,6 +47,12 @@ spec:
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
 {{- end }}
 {{- end }}
diff --git a/operations/helm/charts/mimir-distributed/templates/ruler/ruler-so.yaml b/operations/helm/charts/mimir-distributed/templates/ruler/ruler-so.yaml
@@ -25,20 +25,32 @@ spec:
   triggers:
   - metadata:
       query: max_over_time(sum(sum by (pod) (rate(container_cpu_usage_seconds_total{container="ruler",namespace="{{ .Release.Namespace }}"}[5m])) and max by (pod) (up{container="ruler",namespace="{{ .Release.Namespace }}"}) > 0)[15m:]) * 1000
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       {{- $cpu_request := dig "requests" "cpu" nil .Values.ruler.resources }}
       threshold: {{ mulf (include "mimir.cpuToMilliCPU" (dict "value" $cpu_request)) (divf .Values.ruler.kedaAutoscaling.targetCPUUtilizationPercentage 100) | floor | int64 | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
   - metadata:
       query: max_over_time(sum((sum by (pod) (container_memory_working_set_bytes{container="ruler",namespace="{{ .Release.Namespace }}"}) and max by (pod) (up{container="ruler",namespace="{{ .Release.Namespace }}"}) > 0) or vector(0))[15m:]) + sum(sum by (pod) (max_over_time(kube_pod_container_resource_requests{container="ruler",namespace="{{ .Release.Namespace }}", resource="memory"}[15m])) and max by (pod) (changes(kube_pod_container_status_restarts_total{container="ruler",namespace="{{ .Release.Namespace }}"}[15m]) > 0) and max by (pod) (kube_pod_container_status_last_terminated_reason{container="ruler",namespace="{{ .Release.Namespace }}", reason="OOMKilled"}) or vector(0))
-      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }} 
+      serverAddress: {{ include "mimir.kedaPrometheusAddress" (dict "ctx" $) }}
       {{- $mem_request := dig "requests" "memory" nil .Values.ruler.resources }}
       threshold: {{ mulf (include "mimir.siToBytes" (dict "value" $mem_request)) (divf .Values.ruler.kedaAutoscaling.targetMemoryUtilizationPercentage 100) | floor | int64 | quote }}
       {{- if .Values.kedaAutoscaling.customHeaders }}
       customHeaders: {{ (include "mimir.lib.mapToCSVString" (dict "map" .Values.kedaAutoscaling.customHeaders)) | quote }}
       {{- end }}
+      ignoreNullValues: {{ .Values.kedaAutoscaling.ignoreNullValues }}
+      unsafeSsl: {{ .Values.kedaAutoscaling.ignoreNullValues }}
     type: prometheus
+    {{- if .Values.kedaAutoscaling.authentication.enabled }}
+    authenticationRef:
+      name: {{ include "mimir.resourceName" (dict "ctx" .) }}
+    {{- end }}
 {{- end }}
diff --git a/operations/helm/charts/mimir-distributed/values.yaml b/operations/helm/charts/mimir-distributed/values.yaml
@@ -542,7 +542,21 @@ kedaAutoscaling:
   prometheusAddress: ""
   customHeaders: {}
   pollingInterval: 10
+  ignoreNullValues: true
+  unsafeSsl: false
 
+  # --KEDA trigger authentication settings.
+  # ref: https://keda.sh/docs/2.16/scalers/pulsar/#authentication-parameters
+  authentication:
+    enabled: false
+    authModes: ""
+    secretTargetRef: []
+    # - key: username
+    #   name: my-secret-name
+    #   parameter: username
+    # - key: password
+    #   name: my-secret-name
+    #   parameter: password
 alertmanager:
   enabled: true
   # -- Total number of replicas for the alertmanager across all availability zones