Alice in the world of pods and higher order functions

"We have many reasons to love functional programming. But have you thought that functional programming can be really close to the containers world?
You might have heard about Kubernetes, everyone's buzzing about it. Kubernetes may sound complex especially if you’ve never worked with containers. A surprising thing is that you can actually apply your knowledge of functional programming to understand how Kubernetes works. If you want to know more about the bridge between these two worlds follow me on that path."
Presented at ScalaIO France 2019

Transcript

1. Alice in the world of pods and higher order functions
2. Who am I Big data developer at Captify Women Who Code Kyiv Data Engineering Lead Kubernetes and functional programming fan which loves a good story
3. Alice is an adventurous girl and her curiosity may lead her to epic discoveries of the new worlds…
5. Welcome to the world of pods and higher order functions
6. Functional forest
7. Pods kafka-2 kafka-1 kafka-0 psql-0 psql-1
9. “And these ones?” - asked Alice. “Oh, better don’t look too long at them they are too complex for an unprepared eye. These are called monads”.
14. “Why these houses look like constructor details?” - asks Alice. “This is one of our interesting features. All of our objects constitute type system and they are immutable”
16. CONTROLLERS
17. CONTROLLERS Controller- manager Custom controller
19. kubectl API server Scheduler Controller- manager etcd
20. kubectl API server Scheduler Controller- manager etcd kubelet kube-proxy container Pod Pod container
22. Kubernetes objects map
23. Pod Deployment Replica Set
24. Pod Deployment Replica Set Pod Stateful Set Pod Daemon Set
25. Volume Job Service Pod Deployment Replica Set Pod Stateful Set Pod Daemon Set
26. Volume Job Service Pod Deployment Replica Set Pod Stateful Set Pod Daemon Set CRD
28. IMMUTABLE ??? TYPE SYSTEM
29. TYPE SYSTEM IMMUTABLE DECLARATIVE
30. DECLARATIVE NATURE
31. magic DB
32. pods pure functions monads deployments stateful sets volumes
33. basic-db-9.3.4
34. IMPERATIVE
36. YAML
37. Controller- manager
38. apiVersion: extensions/v1beta1 kind: Deployment metadata: name: magic-db labels: app: magic-db
39. spec: containers: - image: basic-db:9.3.4 name: magic-db env: - name: DB_USER value: wizard - name: DB_NAME value: fairytale
40. kubectl apply -f magic-db.yaml
41. NAME READY STATUS AGE deployment/magic-db 1/1 Running 45s NAME DESIRED CURRENT AGE replicaset/magic-db 1 1 45s
42. NAME READY STATUS AGE pod/magic-db-hg3 0/1 ContainerCreating 4s pod/magic-db-hg3 1/1 Running 45s magic-db- hg3
43. VOLUME
44. apiVersion: v1 kind: PersistentVolumeClaim metadata: name: magic-db-claim labels: app: magic-db
45. spec: accessModes: - ReadWriteOnce resources: requests: storage: 10Gi
46. volumeMounts: - name: magic-db mountPath: /var/lib/magic-db/data volumes: - name: magic-db persistentVolumeClaim: claimName: magic-db-claim
47. NAME STATUS VOLUME CAPACITY AGE magic-db-claim Bound pv-24hjf3 10Gi 32s NAME CAPACITY STATUS CLAIM AGE pv-24hjf3 10Gi Bound magic-db-claim 32s
48. magic-db- hg3
49. SERVICE
50. apiVersion: v1 kind: Service metadata: name: magic-db labels: app: magic-db
51. spec: ports: - containerPort: 5432 name: magic-db
52. NAME TYPE PORT(S) AGE service/magic-db ClusterIP 5432/TCP 5s magic-db- hg3
54. SCALING
55. spec: replicas: 3 template: metadata: labels: app: magic-db
56. NAME READY STATUS AGE pod/magic-db-hg3 1/1 Running 3m pod/magic-db-j3t 1/1 Running 6s pod/magic-db-thw 1/1 Running 4s
57. magic-db- hg3 magic-db- j3t magic-db- thw
58. spec: replicas: 6 template: metadata: labels: app: magic-db
59. NAME READY STATUS AGE pod/magic-db-hg3 1/1 Running 4m13s pod/magic-db-j3t 1/1 Running 53s pod/magic-db-thw 1/1 Running 53s pod/magic-db-yth 1/1 Running 4s pod/magic-db-m1n 1/1 Running 4s pod/magic-db-ito3 1/1 Running 4s
60. magic-db -thw magic-db -j3t magic-db -hg3 magic-db -yth magic-db -m1n magic-db -ito3
61. PURGE
62. kubectl delete -f magic-db.yaml
63. service "magic-db" deleted persistentvolumeclaim "magic-db-claim" deleted deployment “magic-db" deleted
64. Custom controller
65. kubectl API server Custom controller etcd kubelet kube-proxy container CR CR container Operator
67. apiVersion: magic-repo/v1 kind: MagicDatabase metadata: name: magic-db-cluster namespace: default
68. spec: teamID: “Alice-team”
 volume: size: 4Gi numberOfInstances: 3 magic-db: version: “1"
69. users: admin: - Alice databases: magic-world-db: admin
70. NAME READY STATUS AGE pod/magic-db-cluster-0 1/1 Running 3m pod/magic-db-cluster-1 1/1 Running 6s pod/magic-db-cluster-2 1/1 Running 4s magic-db -cluster-2 magic-db -cluster-1 magic-db -cluster-0
71. kubectl exec -it magic-db-cluster-0 /bin/bash @magic-db-cluster-0:/home/magic-db# mdb-sql -U Alice magic-db=#
72. NAME READY STATUS AGE magic-db-operator 1/1 Running 2m33s magic-db-operator-ui 1/1 Running 72s
73. config maps volumes services rbac
74. apiVersion: extensions/v1beta1 kind: Deployment metadata: name: magic-db-operator labels: app: magic-db-operator
75. spec: containers: - image: db-operator: v1.4.0 name: magic-db-operator serviceAccountName: magic-db-operator env: - name: CONFIG_MAP_NAME value: magic-db-operator
77. magic-db- cluster-0
78. kubectl delete magic-db magic-db-cluster
79. KUBERNETES FUNCTIONAL
80. TYPE SYSTEM IMMUTABLE DECLARATIVE DSL
82. magic-db- cluster-0
87. thank you for attention
88. My contact info dead_flowers22 roksolana-d roksolanadiachuk roksolanad