TODO.md

TODO

Vagrant

Create multiple configurations of clusters so people get to see nuances of different distro installs and Kubernetes configurations:

• Ubuntu 18.04
  • Simple
  • Highly Available
• Ubuntu 20.04
  • Simple
  • Highly Available
• CentOS ??
  • Simple
  • Highly Available

Labs

I want these labs to closely mimic the CKA exam - it's quite rigorous proof you know your way around Kubernetes... Here's the current curriculum.

Core Concepts

• Architecture (guided tour of kubeadm cluster)
• Primitives
  • stand up a scaled, stateless web app with ingress and TLS
  • stand up a stateful, clustered database
  • stand up a scheduled batch process

Security

• RBAC Primitives
  • Configure an application to use a unique ServiceAccount
  • Configure a ServiceAccount with read rights on the cluster (!!!)
• Network Policies
  • Lock down a full stack (cache/web/database) with PNP
• Cluster TLS Certificates
  • Rebuild a component's certificates from scratch, preferably with cfssl
• Work With Images Securely
  • Provision a private docker repo in the cluster, stand up Secret for authentication for a Namespace
• Define Security Contexts
  • Configure a privileged Pod and a locked-down Pod (CAPS)
  • Create and use Pod Security Policies
• Secure key-value store
  • Encrypt etcd and validate

Installation, Configuration, and Validation

• Simple
  • kubeadm install for masters, nodes, CNI, secure communications, master e2e tests, node e2e tests
• Highly Available
  • kubeadm install for masters, nodes, CNI, secure communications, master e2e tests, node e2e tests

Cluster Maintenance

• Cluster Upgrades
  • Simple Upgrade
  • Highly Available Upgrade
• OS Upgrades
• Backup and Restore
  • Simple
    • Unencrypted etcd backup/restore
    • Encrypted etcd backup/restore
  • Highly Available
    • Unencrypted etcd backup/restore
    • Encrypted etcd backup/restore

Networking

• Cluster networking
  • CNI configuration
  • Understand node network
  • Cluster DNS
• Service networking
  • Deploy network load balancer
• Ingress rules

Troubleshooting

• Application failure
• Control plane failure
• Worker node failure
• Network failure

Application Life Cycle Management

• Deployments
  • Updates, rollbacks, scaling
• Jobs
  • Scaling, CronJobs
• StatefulSets
  • Updates, rollbacks, scaling, caveats
• Pods
  • ConfigMap/Secret usage
  • Self-Healing applications

Storage

• Cluster storage
  • PersistentVolumes, StorageClasses, PersistentVolumeClaims
• Pod storage
  • Volumes (access modes), configuration through Volumes

Scheduling

• Scheduling Pods
  • NodeSelector, label selectors, static pods, schedule events, resource limits
• DaemonSets
• kube-scheduler
  • Multiple schedulers, configure scheduler, events

Logging/Monitoring

• Cluster components
  • Monitor / manage logs
• Applications
  • Monitor / manage logs