Containers and Cloud Native overview
67 topics in 13 lessons, each in one line. Every topic opens the page that teaches it.
After this course you can
- Explain what a container is, build a small image and run several containers together.
- Read and write the Kubernetes objects that run, expose, configure and store an application.
- Find out why a workload is not running, and fix it with kubectl, the way the hands-on exams expect.
1Why containers, and running your first one
2Running containers
- PortsA container's ports are private until you publish them
- VolumesWhat a container writes dies with it, unless it is in a volume
- EnvironmentOne image runs everywhere; the settings come from outside
- NetworksOn a shared network, containers find each other by name
- ComposeCompose writes the whole picture down in one file
3Building and sharing images
4Why Kubernetes, and what is in a cluster
- Why KubernetesYou state what should run; the cluster keeps it true
- Desired stateWatch, compare, act: the loop that runs everything
- Inside a clusterA control plane decides; worker nodes run the containers
- One requestOne kubectl apply, followed through the cluster
- ManifestsEvery object has the same shape: you write the spec
5Pods and Deployments
6Services and networking
- ServicesA Service is one stable address in front of changing pods
- Service typesThree Service types, each built on the one before
- DNSEvery Service has a name; programs use names, not addresses
- IngressAn Ingress shares one entrance among many Services
- NetworkPolicyEverything can reach everything, until a policy says no
7Configuration and secrets
- ConfigMaps and SecretsSettings are objects in the cluster, handed to pods
- The pod manifestA manifest names its settings; it does not contain them
- SecretsA Secret is not secret by itself
- Requests and limitsA request is a promise; a limit is a ceiling
- Quality of serviceWhen a node runs out, the pods that promised least go first
8Storage
9Scheduling and scaling
- The schedulerFirst rule nodes out, then pick the best of the rest
- PlacementFive pods, three nodes, and the reason for each
- SteeringAffinity draws pods in; a taint pushes them away
- Scaling podsThe autoscaler holds load at a target by changing the count
- Scaling nodesWhen pods no longer fit, the cluster itself grows
10Security in the cluster
11Packaging and delivery
- The problemOne copy of what is the same; write down only what differs
- HelmA chart is manifests with blanks, and values to fill them
- KustomizeKustomize lays small patches over plain manifests
- GitOpsGit holds the truth; the cluster pulls itself into line
- The landscapeKubernetes is one project among many, each with one job
12Operating and fixing a cluster
13The pod that keeps restartingWorked case
- The problemPackage once, and let the cluster keep it running
- The first deploymentA Deployment keeps pods running; a Service gives them one address
- Version 2The rollout stalls safely while version 2 restarts again and again
- The causeKilled for using more memory than its limit
- The fixSize the limit from the measured peak, and give a slow start its own probe
- The trade-offEvery megabyte requested is one no other pod can use
- The finished pictureA healthy workload is six small, measured decisions
- A real caseAt Monzo in 2017, a stuck rollout became a full outage
Sources and credit
49 sourcesThe chapters and drawings are written by NeuroqAI from these sources. Each lesson lists its own at the end of its chapter, and figures on the pages are taken from them or computed from them.