Azure Kubernetes Service

This integration can be enabled via the Connect an Azure Account page.

This integration has 16 metrics.

List of Alerts

List of Dashboards

Azure Kubernetes Service

The dashboard provides information on the Azure Azure Kubernetes Service (AKS).

List of Metrics

Monitoring and Troubleshooting Azure Kubernetes Service

This document describes important metrics and queries that you can use to monitor and troubleshoot Azure Kubernetes Service.

Cluster State

Nodes not ready

Use the following query to check if there are nodes with NotReady state:

azure_containerservice_managedclusters_kube_node_status_condition_avg{condition="Ready", status2="NotReady"}

A return value other than 0 indicates a problem in the cluster.

CPU cores

Use the following query to get the number of cores on your cluster:

azure_containerservice_managedclusters_kube_node_status_allocatable_cpu_cores_avg

Memory Available

Use the following query to get the available memory:

azure_containerservice_managedclusters_kube_node_status_allocatable_memory_bytes_avg

A low value can indicate memory pressure, a new node should be created manually or by the autoscaler.

Cluster nodes

CPU usage

Use the following query to get the CPU percentage usage:

azure_containerservice_managedclusters_node_cpu_usage_percentage_avg

A return value greater than 95 indicates CPU pressure in a node.

Memory RSS

Memory RSS is the amount of anonymous and swap cache memory.

Use the following query to get the CPU percentage usage:

azure_containerservice_managedclusters_node_memory_rss_percentage_avg

Memory working set

The memory working set includes the amount of kernel memory, dirty memory, and recently accessed memory.

Use the following query to get the CPU percentage usage:

azure_containerservice_managedclusters_node_memory_working_set_percentage_avg

A return value greater than 95 indicates memory pressure in a node.

Disk usage

Use the following query to get the disk usage:

azure_containerservice_managedclusters_node_disk_usage_percentage_avg

A high value return indicates that a disk is running low space.

Network In

Use the following query to get network bandwidth in:

azure_containerservice_managedclusters_node_network_in_bytes_avg

Network Out

Use the following query to get network bandwidth out:

azure_containerservice_managedclusters_node_network_out_bytes_avg

API Inflight Requests

Use the following query to get the API requests:

azure_containerservice_managedclusters_apiserver_current_inflight_requests_avg