csp-mixin(gcp compute engine): Add alert for CPU (#1345)

* csp-mixin(gcp compute engine): Add alert for CPU

* split alerts in multiple files

* importRules directly on mixing

* use config prometheus for each provider

* fix lint

* fix alerts

* remove alerts libsonnet file

---------

Co-authored-by: yduartep <[email protected]>

Author: anaivanov

csp-mixin/.lint

@@ -71,3 +71,9 @@ exclusions:
   panel-title-description-rule:
     entries:
     - dashboard: Azure Service Bus
+    - dashboard: GCP Compute Engine
+    - dashboard: Azure Virtual Machines
+  target-rate-interval-rule:
+    entries:
+    - dashboard: GCP Compute Engine
+

csp-mixin/alerts.libsonnet

@@ -0,0 +1,30 @@
+groups:
+  - name: azure
+    rules:
+      - alert: AzureVMHighCpuUtilization
+        expr: |
+          avg by (job,resourceGroup,subscriptionName,resourceName) (azure_microsoft_compute_virtualmachines_percentage_cpu_average_percent{job=~".+",resourceGroup=~".+",subscriptionName=~".+",resourceName=~".+"})  > 85
+        for: 5m
+        keep_firing_for: 10m
+        labels:
+          severity: critical
+          service: 'Azure Virtual Machines'
+          namespace: cloud-provider-azure
+        annotations:
+          summary: 'CPU utilization is too high.'
+          description: 'The VM {{ $labels.resourceName }} is under heavy load and may become unresponsive.'
+          dashboard_uid: '58f33c50e66c911b0ad8a25aa438a96e'
+
+      - alert: AzureVMUnavailable
+        expr: |
+          avg by (job,resourceGroup,subscriptionName,resourceName) (azure_microsoft_compute_virtualmachines_vmavailabilitymetric_average_count{job=~".+",resourceGroup=~".+",subscriptionName=~".+",resourceName=~".+"}) != 1
+        for: 5m
+        keep_firing_for: 10m
+        labels:
+          severity: critical
+          service: 'Azure Virtual Machines.'
+          namespace: cloud-provider-azure
+        annotations:
+          summary: 'VM unavailable.'
+          description: 'The VM {{ $labels.resourceName }} is not functioning or crashed, which may require immediate action.'
+          dashboard_uid: '58f33c50e66c911b0ad8a25aa438a96e'

csp-mixin/alerts/azure-alerts.yml

@@ -0,0 +1,16 @@
+groups:
+  - name: gcp
+    rules:
+      - alert: GcpCEHighCpuUtilization
+        expr: |
+          100 * avg by (job,project_id,instance_name) (stackdriver_gce_instance_compute_googleapis_com_instance_cpu_utilization{job=~".+",project_id=~".+",instance_name=~".+"}) > 85
+        for: 5m
+        keep_firing_for: 10m
+        labels:
+          severity: critical
+          service: 'Compute Engine'
+          namespace: cloud-provider-gcp
+        annotations:
+          summary: 'CPU utilization is too high.'
+          description: 'The VM {{ $labels.instance_name }} is under heavy load and may become unresponsive.'
+          dashboard_uid: 'f115fe73641347c43415535d77e2dc0f'

csp-mixin/alerts/gcp-alerts.yml

@@ -19,5 +19,13 @@
     groupLabels: ['job', 'resourceGroup', 'subscriptionName'],
     instanceLabels: ['resourceName'],
     metricsSource: 'azuremonitor',
+
+    local importRules(rules) = {
+      groups+: std.parseYaml(rules).groups,
+    },
+
+    prometheus: {
+      alerts: importRules(importstr 'alerts/azure-alerts.yml'),
+    },
   },
 }

csp-mixin/azureconfig.libsonnet

@@ -19,5 +19,12 @@
     groupLabels: ['job'],
     instanceLabels: ['bucket_name'],
     metricsSource: 'stackdriver',
+    local importRules(rules) = {
+      groups+: std.parseYaml(rules).groups,
+    },
+
+    prometheus: {
+      alerts: importRules(importstr 'alerts/gcp-alerts.yml'),
+    },
   },
 }

csp-mixin/gcpconfig.libsonnet

@@ -15,7 +15,7 @@ local commonlib = import 'common-lib/common/main.libsonnet';
       dashboards: (import './dashboards.libsonnet').new(this),
     },
     prometheus: {
-      alerts: (import './alerts.libsonnet').new(this),
+      alerts: this.config.prometheus.alerts,
       recordingRules: {},
     },
     asMonitoringMixin(): {