# you can disable an alarm notification by setting the 'to' line to: silent

# --- API Server ---

 template: am_aks_apiserver_cpu
       on: azure_monitor.aks.apiserver_cpu
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (75) : (80))
     crit: $this > (($status == $CRITICAL) ? (85) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS API server CPU on ${label:resource_name}
     info: Average API server CPU utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

 template: am_aks_apiserver_memory
       on: azure_monitor.aks.apiserver_memory
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (75) : (80))
     crit: $this > (($status == $CRITICAL) ? (85) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS API server memory on ${label:resource_name}
     info: Average API server memory utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

 template: am_aks_apiserver_inflight_requests
       on: azure_monitor.aks.apiserver_inflight_requests
    class: Workload
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: requests
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (400) : (600))
     crit: $this > (($status == $CRITICAL) ? (600) : (800))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS API server inflight requests on ${label:resource_name}
     info: Average number of inflight requests to the API server on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# --- etcd ---

 template: am_aks_etcd_cpu
       on: azure_monitor.aks.etcd_cpu
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (75) : (80))
     crit: $this > (($status == $CRITICAL) ? (85) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS etcd CPU on ${label:resource_name}
     info: Average etcd CPU utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

 template: am_aks_etcd_memory
       on: azure_monitor.aks.etcd_memory
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (75) : (80))
     crit: $this > (($status == $CRITICAL) ? (85) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS etcd memory on ${label:resource_name}
     info: Average etcd memory utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

 template: am_aks_etcd_database
       on: azure_monitor.aks.etcd_database
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (75) : (80))
     crit: $this > (($status == $CRITICAL) ? (85) : (90))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS etcd database usage on ${label:resource_name}
     info: Average etcd database utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           High etcd database usage can lead to cluster instability.
       to: sysadmin

# --- Autoscaler ---

 template: am_aks_autoscaler_safe_to_autoscale
       on: azure_monitor.aks.autoscaler_health
    class: Availability
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of safe_to_autoscale
    units: state
    every: 1m
     warn: $this != nan AND $this < 1
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS autoscaler unsafe on ${label:resource_name}
     info: Cluster autoscaler reports the cluster is not safe to autoscale on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

 template: am_aks_autoscaler_unschedulable_pods
       on: azure_monitor.aks.autoscaler_unschedulable_pods
    class: Errors
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: pods
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (0) : (5))
     crit: $this > (($status == $CRITICAL) ? (5) : (20))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS unschedulable pods on ${label:resource_name}
     info: Number of pods that cannot be scheduled by the cluster autoscaler on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Indicates insufficient cluster capacity.
       to: sysadmin

# --- Node CPU ---

 template: am_aks_node_cpu
       on: azure_monitor.aks.node_cpu_percentage
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (80) : (90))
     crit: $this > (($status == $CRITICAL) ? (90) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS node CPU on ${label:resource_name}
     info: Average node CPU utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# --- Node Memory ---

 template: am_aks_node_memory_working_set
       on: azure_monitor.aks.node_memory_working_set_percentage
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (80) : (85))
     crit: $this > (($status == $CRITICAL) ? (85) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS node memory working set on ${label:resource_name}
     info: Average node memory working set utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

 template: am_aks_node_memory_rss
       on: azure_monitor.aks.node_memory_rss_percentage
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (80) : (85))
     crit: $this > (($status == $CRITICAL) ? (85) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS node memory RSS on ${label:resource_name}
     info: Average node memory RSS utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# --- Node Disk ---

 template: am_aks_node_disk
       on: azure_monitor.aks.node_disk_percentage
    class: Utilization
     type: Kubernetes
component: AKS
   lookup: average -5m unaligned of average
    units: percentage
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (75) : (80))
     crit: $this > (($status == $CRITICAL) ? (85) : (90))
    delay: down 5m multiplier 1.5 max 1h
  summary: AKS node disk usage on ${label:resource_name}
     info: Average node disk utilization on AKS cluster ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           High disk usage can cause pod evictions.
       to: sysadmin