Practical Kubernetes Guide

Explore how networking plugins shape Kubernetes networks in Azure Kubernetes Service and how pod IPs are assigned. Compare Kubernetes default, Azure CNI, overlay mode, and bring-your-own CNI plugins.

Explore kubenet as the default network mode in azure kubernetes service, where nodes borrow subnet ips while pods receive a pod cidr, with nat routing and limited route table entries.

Explore azure cni network mode, where each pod receives an ip address from the subnet and requires ip planning to avoid exhaustion, while clarifying pod vs service cidr behavior.

Explore how Azure CNI overlay mode solves IP exhaustion by assigning each node a unique pod CIDR, isolating pod IPs from the subnet and avoiding extra route hops.

Compare Kubernetes with Azure CNI overlay, highlighting up to 1000 nodes per cluster, no per-node route-table limits, and simplified pod networking.

Learn to fetch and analyze AKS control plane logs, including the API server, scheduler, and controller manager, by configuring diagnostic settings and sending data to log analytics.

Connect to your Azure Kubernetes cluster with kubelogin and download credentials. Use the Cube plugin for non-interactive authentication with service principal, managed identity, or workload identity in pipelines.

Explore manual and automatic scaling in Kubernetes with the horizontal pod autoscaler, which scales pod replicas based on CPU and memory utilization within deployment minimum and maximum replicas.

Scale a Kubernetes cluster with AKS by manually adjusting node counts via the Azure portal or CLI. Compare manual scaling with autoscaling options, including the HPA, based on CPU utilization.

Scale a Kubernetes cluster with the cluster autoscaler in AKS, configuring min and max nodes to meet demand. See how pending pods trigger VM provisioning and scaling with HPA.

Explore how Kubernetes operators automate complex application management by extending the API with controllers and CRDs, and see examples like Elasticsearch and Istio.

Install and configure Prometheus and Grafana operators in Kubernetes using the operator lifecycle manager and Kube Prometheus. Explore CRDs, dashboards, and pre-configured monitoring components.

Explore deploying a Kubernetes logging stack using fluentd, fluent bit, Elasticsearch, and Kibana with operators. Learn to install the logging and Elasticsearch operators and access Kibana for logs.

Explore how Kubernetes service accounts grant machine-to-machine access to the API server via roles and role bindings, enabling pods to access resources such as deployments, services, ingress, configmaps, and secrets.

Learn how to use Kubernetes role and role binding to grant tenants access to their own namespaces, enabling deployment of pods, deployments, secrets, and configmaps; explore cluster-wide roles for administrators.

Create a pod reader role and a role binding in a namespace, assign them to users and groups, and verify access with kubectl can-i for authorized and unauthorized users.

Use workload identity with OpenID Connect to access Azure resources passwordlessly from Kubernetes. Bind a service account to an Azure managed identity to obtain access tokens and securely reach resources.

Explore passwordless identities in a Kubernetes workflow by linking a service account to an Azure managed identity, granting access to Key Vault secrets via workload identity.