Monitor Azure Kubernetes Service (AKS)

• An active Azure subscription.
• An Azure Kubernetes Service (AKS) cluster deployed. If you don't have one, follow the quickstart guide to create an AKS cluster.
• The Azure CLI installed and configured.
• kubectl command-line tool installed and configured to connect to your AKS cluster.

To enable Container Insights for your AKS cluster, you can use the Azure CLI:

az aks enable-addons -a monitoring -n MyAKSCluster -g MyResourceGroup
                            

Replace MyAKSCluster and MyResourceGroup with your AKS cluster name and resource group name, respectively.

This command deploys the Container Insights agent to your cluster and configures it to send metrics and logs to your Azure Monitor workspace.

1. Navigate to your AKS cluster in the Azure portal.
2. In the left-hand menu, under "Monitoring", select "Insights".
3. You will see an overview dashboard with key metrics like CPU usage, memory usage, network traffic, and more.
4. Explore the "Containers" section to view node performance, pod health, and running applications.
5. Use the "Live Data" view to see real-time logs and events from your cluster.

Container Insights offers pre-built workbooks that visualize cluster health, performance, and capacity. You can customize these workbooks or create your own to meet specific needs.

You can set up alert rules in Azure Monitor based on various metrics and log queries:

1. Go to your AKS cluster and select "Monitoring" > "Alerts".
2. Click on "Create alert rule".
3. Select a scope (e.g., your AKS cluster resource).
4. Define the condition: Choose a signal (metric or log), set thresholds, and aggregation granularity. For example, you could alert if CPU utilization for a deployment exceeds 80% for 5 minutes.
5. Choose an action group: This determines what happens when an alert is triggered (e.g., send an email, trigger a webhook, run an Azure Function).
6. Give the alert rule a name and description, then save it.

To query logs directly:

1. Navigate to your Log Analytics workspace associated with your AKS cluster.
2. Select "Logs".
3. Use Kusto Query Language (KQL) to query your cluster data. Some common KQL queries for AKS monitoring include:
   • KubePodInventory | summarize count() by PodStatus (Count pods by status)
   • KubeNodeInventory | where PowerState != "running" (Find nodes not running)
   • KubeEvents | where Type = "Error" (View error events)

You can also create custom dashboards in Azure Monitor to visualize specific log query results.

• Application Performance Monitoring (APM): Integrate Application Insights with your applications running on AKS for detailed tracing, dependency mapping, and performance diagnostics.
• Service Meshes: For microservice architectures, consider using a service mesh like Istio or Linkerd, which can provide advanced traffic management, security, and observability features.
• Prometheus and Grafana: While Azure Monitor provides excellent integrated solutions, you can also deploy Prometheus for metric collection and Grafana for dashboarding directly within your AKS cluster for highly customized monitoring.
• Security Monitoring: Leverage Azure Security Center and Azure Sentinel for comprehensive security posture management and threat detection on your AKS clusters.