Skip to content

Instantly share code, notes, and snippets.

@invalidred

invalidred/intro-to-kubernetes.md

Created January 17, 2019 00:51
Show Gist options
  • Save invalidred/ef949003d6eda0b41b8b0f2feb5990d8 to your computer and use it in GitHub Desktop.
Save invalidred/ef949003d6eda0b41b8b0f2feb5990d8 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
intro-to-kubernetes.md

Chapter 1: Introduction To Kubernetes

Summary

Moving from Monoliths to micro-services which allows components to be developed, deployed, updated and scaled independently. This allows to develop rapidly based on business needs.

However with more deployable components it is increasingly harder to configure, manage and keep the whole system running. We need automation which includes scheduling of those components into our servers, auto configuration, supervision and failure-handling. This is how Kubernetes can help.

Kubernetes allows devs to deploy apps without the help of operations (ops) team. It also helps ops by automatically scheduling and monitoring those apps in event of hardware failure.

::Kubernetes abstracts the entire data-centre as a single enormous computational resource::. It allows you to deploy components without knowing the infrastructure. You give your component(s) and Kubernetes will select the best server for your app.

Core Idea

The system is composed of a master node and any number of worker nodes. When you deploy your apps to master Kubernetes deploys them to any number of worker nodes. What node a component lands on doesn’t matter to dev or system admin.

[image:0A3ED9ED-C762-4448-B7F4-281769737540-814-00007CA2E277B2DC/Screen Shot 2019-01-16 at 7.22.33 PM.png]

Dev can specify the apps that need to be together and Kubernetes will deploy them on same worker nodes and other apps. Other apps will spread across the cluster, however they can talk to each other the same way regardless of where they are deployed.

Basic Services Kubernetes Provides

Kubernetes provides a lot of services, enabling the developer to focus on code and not worry about integrating with infrasctrue

  • scaling
  • service discovery
  • load-balancing
  • self-healing
  • leader election

Kubernetes Architecture in Depth

  • It consists of master node which contains Kubernetes Control Plane that manages Kubernetes system.
  • It also consists of worker nodes

[image:C08D46BD-68D2-463A-A54A-5F2DEB783314-814-00007D61B37895EA/078AD92B-96BD-40C3-AE3C-40B3218FC28E.png]

The Control Plane

It control the cluster and can be hosted on one machine or multiple machines for higher availability

  • API Server you can other control plane components communicate with
  • Scheduler helps schedule your app by binding it to a worker node
  • Controller Manager does cluster level functions such as component replication, tracking worker nodes, handing failure etc..
  • etcdreliable distributed data store to store cluster config

The (worker) Nodes

These are Nodes that run your containerized components.

  • You can use Docker or any Open Container Compliant technology such as rkt
  • Kublettalks to API server and manages containers on it node
  • kube-proxy load-balances network traffic amongst app components

#kb/books/kubernetes_in_action/Chapter_1

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment