Slide 1

Slide 1 text

@BastianHofmann Creating a fast Kubernetes Development Workflow Bastian Hofmann

Slide 2

Slide 2 text

No content

Slide 3

Slide 3 text

Container orchestration platform

Slide 4

Slide 4 text

Deploy, run and scale your services in isolated containers

Slide 5

Slide 5 text

No vendor lock in

Slide 6

Slide 6 text

Runs on

Slide 7

Slide 7 text

Your laptop

Slide 8

Slide 8 text

Bare metal

Slide 9

Slide 9 text

Cloud Providers

Slide 10

Slide 10 text

And if you don't want to install and maintain Kubernetes yourself

Slide 11

Slide 11 text

Managed Kubernetes

Slide 12

Slide 12 text

No content

Slide 13

Slide 13 text

Standardized APIs

Slide 14

Slide 14 text

It works the same everywhere*

Slide 15

Slide 15 text

It works the same everywhere* *mostly

Slide 16

Slide 16 text

This talk is about how to use Kubernetes

Slide 17

Slide 17 text

Not only for production workloads

Slide 18

Slide 18 text

But in your development workflows

Slide 19

Slide 19 text

Goal: Development environment as close to production as possible

Slide 20

Slide 20 text

Kubernetes' standardized API makes this easier

Slide 21

Slide 21 text

Agenda

Slide 22

Slide 22 text

Deployment of a micro-service application

Slide 23

Slide 23 text

Some tools to help with local development of this application on Kubernetes

Slide 24

Slide 24 text

Let's have a look at the sample application

Slide 25

Slide 25 text

No content

Slide 26

Slide 26 text

OpenStack Cloud LoadBalancer NGINX Ingress Controller NGINX Ingress Controller NGINX Ingress Controller web-application web-application MySQL Primary MySQL Secondary quote-svc quote-svc hello-svc hello-svc

Slide 27

Slide 27 text

external-dns to create DNS entries automatically

Slide 28

Slide 28 text

cert-manager to retrieve Let's Encrypt certificates automatically

Slide 29

Slide 29 text

Database is managed by an Operator

Slide 30

Slide 30 text

MySQL Operator MySQLCluster MySQL pods MySQL statefulset Kubernetes controller manager Discovers Creates Creates Discovers Monitors and manages

Slide 31

Slide 31 text

If you are interested in the code and how to set it up: https:/ /github.com/syseleven/ golem-workshop

Slide 32

Slide 32 text

Demo

Slide 33

Slide 33 text

Writing this YAML files is tedious

Slide 34

Slide 34 text

YAML files are tied to a specific version and a specific environment

Slide 35

Slide 35 text

Production

Slide 36

Slide 36 text

Staging

Slide 37

Slide 37 text

Development

Slide 38

Slide 38 text

Per Development team

Slide 39

Slide 39 text

Per branch

Slide 40

Slide 40 text

Per developer

Slide 41

Slide 41 text

We need to maintain multiple, very similar YAML files with slightly different versions and configuration

Slide 42

Slide 42 text

"Templating"

Slide 43

Slide 43 text

Great tools because of standardized Kubernetes API

Slide 44

Slide 44 text

Helm

Slide 45

Slide 45 text

No content

Slide 46

Slide 46 text

Allows to install applications

Slide 47

Slide 47 text

So called "charts"

Slide 48

Slide 48 text

$ helm install stable/wordpress \ --name my-blog \ --namespace blog

Slide 49

Slide 49 text

Charts can depend on other charts

Slide 50

Slide 50 text

Multiple deployments of one chart possible

Slide 51

Slide 51 text

Different release names

Slide 52

Slide 52 text

Different namespaces

Slide 53

Slide 53 text

Configuration with values

Slide 54

Slide 54 text

No content

Slide 55

Slide 55 text

$ helm install stable/wordpress \ --name my-blog \ --namespace blog \ -f my-config-values.yaml

Slide 56

Slide 56 text

Demo

Slide 57

Slide 57 text

Writing your own charts is fairly easy

Slide 58

Slide 58 text

Scaffolding to get started

Slide 59

Slide 59 text

$ helm create quote-svc

Slide 60

Slide 60 text

Helm lint

Slide 61

Slide 61 text

Helm kubeval

Slide 62

Slide 62 text

Helm test

Slide 63

Slide 63 text

Demo

Slide 64

Slide 64 text

Alternatives: Kustomize

Slide 65

Slide 65 text

This works now great for production or staging or CI

Slide 66

Slide 66 text

Still, for development:

Slide 67

Slide 67 text

Make a code change

Slide 68

Slide 68 text

Build docker image

Slide 69

Slide 69 text

Push docker image

Slide 70

Slide 70 text

Run helm install/upgrade with new image version

Slide 71

Slide 71 text

Can this be quicker?

Slide 72

Slide 72 text

Run everything locally

Slide 73

Slide 73 text

docker-compose

Slide 74

Slide 74 text

Duplication of the definition of how to run a container

Slide 75

Slide 75 text

Inconsistencies

Slide 76

Slide 76 text

If you have a lot of services, you have to run a lot locally

Slide 77

Slide 77 text

Some services locally, some remote

Slide 78

Slide 78 text

Service Discovery

Slide 79

Slide 79 text

Not every service is exposed to the Internet

Slide 80

Slide 80 text

Shared resources with other developers?

Slide 81

Slide 81 text

Other options?

Slide 82

Slide 82 text

Tilt

Slide 83

Slide 83 text

$ tilt up

Slide 84

Slide 84 text

Watches for code changes

Slide 85

Slide 85 text

Rebuilds docker image

Slide 86

Slide 86 text

Deploys to Kubernetes

Slide 87

Slide 87 text

Sets up port-forwarding

Slide 88

Slide 88 text

Can sync changed files directly into a running container

Slide 89

Slide 89 text

Demo

Slide 90

Slide 90 text

Alternatives: Skaffold Garden

Slide 91

Slide 91 text

Debugging containers

Slide 92

Slide 92 text

Most containers do not have all the debugging tools included

Slide 93

Slide 93 text

Kubectl debug

Slide 94

Slide 94 text

Debugging network traffic between containers

Slide 95

Slide 95 text

Kubectl sniff

Slide 96

Slide 96 text

Demo

Slide 97

Slide 97 text

Another approach

Slide 98

Slide 98 text

No content

Slide 99

Slide 99 text

Creates a two-way proxy between the Kubernetes cluster and you

Slide 100

Slide 100 text

$ telepresence T: Starting proxy with method 'vpn-tcp'... @fhgbvx65xg|bash-3.2$ curl http://quote-svc/quote | jq '.' [ { "ID": 503, "title": "stefan sagmeister", "content": "

...

\n", "link": "https://quotesondesign.com/stefan- sagmeister-2/" } ]

Slide 101

Slide 101 text

Swap a running deployment in the cluster with a local process

Slide 102

Slide 102 text

... or a locally running docker container

Slide 103

Slide 103 text

$ telepresence --swap-deployment quote-svc --namespace dev-flow-demo --expose 3000 --run npm run debug T: Starting proxy with method 'vpn-tcp',... T: Forwarding remote port 3000 to local port 3000.... > [email protected] debug /Users/bhofmann/forge_test/quote- svc > nodemon --inspect quote-svc.js [nodemon] watching: *.* [nodemon] starting `node --inspect quote-svc.js` Debugger listening on ws://127.0.0.1:9229/83aa27ac- d879-4b50-a228-440354cca791 quote svc listening on port 3000!

Slide 104

Slide 104 text

Demo

Slide 105

Slide 105 text

Summary

Slide 106

Slide 106 text

Powerful

Slide 107

Slide 107 text

Great tooling because of common APIs

Slide 108

Slide 108 text

Especially great if you have multiple services and don't want to run everything locally

Slide 109

Slide 109 text

Test it 30 days For free Visit us at our booth

Slide 110

Slide 110 text

[email protected] https:/ /twitter.com/BastianHofmann http:/ /speakerdeck.com/u/bastianhofmann https:/ /github.com/syseleven/golem-workshop