Nod –> VM or phisical Machine

• Pod –> smallest component (one container per pod usually, but can be more) each pod gets an IP (internal, dynamic – K8s virt.net)
• Service –> Fix IP for Pod (internal or external) act as loadbalancer also, use less busy nod
• Ingress –> Entry point for end user
• ConfigMap –> External conf for apps. URLs, IP, ports, credentials
• Secret –> for credentials. base64 encoded
• Volumes –> external phisical storage. Localhost, Cloud, NSF
• Deployment –> blueprints for app pod, how many replicas…
• StatefulSet –> for DB replica sinc (DB are often hosted outside claster)

• basic Architecture….
• 2 tipes of nodes. Master and Slave
• 3 process must run on every worker (slave) nod:
• 1. Container runtime, docker or something else
• 2. Kublet – Kubernati process
• 3. Kube proxy – forwards request
• Master nod runs 4 processes:
• 1. API server: cluster gateway, authentication for deploying..client access, UI, cubectl, cubernati dashboard…etc
• 2. Scheduler… decide on which nod to put the pod (less used)
• 3. Controller manager…detects pod crashing and restarts them (cluster state changes)
• 4. ETCD key value store… cluster state info

minikube – for testing. master and worker nod on single machine.

minikube start –vm-driver=hyperkit	or docker, Hyper-V,KVM, Parallels, Virtual Box, VMware
kubectl get nodes	get status of nodes
minikube status	—– | | —–
kubectl version	display version
kubectl get pod (-o wide)	check pods (wide)
kubectl get services	check services
kubectl create deployment NAME –image=image	create deployment with pod in it
kubectl get deployment [name] (-o yaml)	check depooyment (3rd part, config status)
kubectl get replicaset	show replicas
kubectl edit deployment [name]	edit config file
kubectl logs [pod name]	for debagging
kubectl describe pod [pod name]	for debagging, more info
kubectl exec -it [pod_name] — bin/bash	get pod terminal
kubectl delete deployment [name]	rm pod
kubectl apply/delete -f [file_name]	create or delete deployment from config file (.yaml)

### config.yaml ### example ###
### eatch file is made of 3 parts ###
### metadata: ...names
### specification: ...any kind of config. attributs will be specific to the kind
### status: ...will be automatically generated by kubernates
###++++++++++++++++####
#### yaml file iz very strict about indentation - sintax ###

### tamplate: has its own metadata and spec. apply to pod ###
### Connection is established using labels and selectors ###

apiVersion: apps/v1
kind: deployment
metadata:
  name: nginx-dpl
  labels:
    app: nginx
spec:           ##### spec for deployment
  replicas: 1
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:        ##### spec for pod
      containers:
      - name: nginx
        image: nginx:1.16
        ports:
        - containerPort: 80

Set Secret

### --- separate config (multiple conf in one file possible) ###
### deployment and service yaml goes in one file ###

### ConfigMap... external conf ###

apiVersion: v1
kind: ConfigMap
metadata:
  name: mongodb-comfigmap
data:
  databese_url: mongodb-service   ### name of the service is URL

---
#### way to reference it..###
    
     env:
     - name: ME_CONFIG_MONGODB_SERVER
       valueFrom:
         configMapKeyRef:
           name: mongodb-configmap
           key: database_url

### externenal service for mongo UI ####

apiVersion: v1
kind: service
metadata:
  name: mongo-express-service
spec:
  selector:
    app: mongo-express
  type: LoadBalancer      ### this make it external
  ports:
    - protocol: TCP
      port: 8081
      targetPort: 8081
      nodePort: 30000     ### port for external IP you need to put in browser, 30000-32767

minikube service [name of th service]	to get external IP
kubectl cluster-info	show cluster info

Namespaces ### virtual cluster inside cluster, organise resources

cubectl get namespace    ### 4 default namespaces

default              ### resources you make, goes here
kube-node-lease      ### each node has associated lease
kube-public          ### publicely accessible data
kube-system          ### do NOT modify, for system use only
kubernates-dashboard ### minikube only

######################################
config file
#############

apiVersion: v1
kind: ConfigMap
metadata:
  name: mysql-cfgmap
  namespace: my-namespace
data: 
  db_url: mysql-service.database ### database is namespace

###############################
# each namespace must define own ConfigMap
# same for secret
# service can be shared beetween namespaces
# volumes and nod can not be namespaced

kubectl create namespace [namespace_name]	create new namespace
kubectl get configmap -n [my_namespace]	show configmap of specific namespace
cubens [my_namespace]	change default ns, cubectx has to be installed first.

Ingress – external access via domain (no ports)

apiVersion: v1
kind: service
metadata:
  name: myapp-ext-service
spec:
  selector:
    app: myapp
  type: LoadBalancer
  ports:
    - protocol: TCP
      port: 8080
      targetPort:8080
      nodePort: 35010 ### public IP port

########################################

apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadate:
  name: myApp-ingress
spec:
  rules:
  - host: myapp.com   ### entrypoint for end user
    http:             ### not browser http
      paths:    ### routing to internal service, url path
      - backend:
          serviceName: myapp-ext-service
          servicePort: 8080      ### internal service port

Ingress controler, another pod that must be installed on nod to make ingress works. it manages redirections. k8s nginx ingress controler is from kubernates, but there are others.

minikube addons enabled ingress	starts and configure nginx ingress controler in minikube
kubectl get pod -n kube-system	check if it is running

multiple paths with the same host

apiVersion: networking.k8s.io:v1beta1
kind: Ingress
metadata:
  name: multi_path
  annotations:
    nginx.ingress.kubernates.io/rewrite-target: /
spec:
  rules:
  - host: myapp.com
    http:
      paths:
      - path: /analytics
        backend:
          serviceName: analytics-service
          servicePort: 3000
      - path: /shopping
        backend:
          serviceName: shopping-service
          servicePort: 5000

###########################
### or with subdomains ###
##########################

apiVersion: networking.k8s.io:v1beta1
kind: Ingress
metadata:
  name: subdomains
  annotations:
    nginx.ingress.kubernates.io/rewrite-target: /
spec:
  rules:
  - host: analytics.myapp.com
    http:
      paths:
        backend:
          serviceName: analytics-service
          servicePort: 3000
  - host: shopping.myapp.com
    http:
      paths:
        backend:
          serviceName: shopping-service
          servicePort: 5000

TLS Cert

apiVersion: networking.k8s.io:v1beta1
kind: Ingress
metadata:
  name: tls-cert-example
spec:
  tls:
    - hosts:
      - myapp.com
    seceretName: myapp-tls-cert
  rules:
    - host: myapp.com
      http:
        paths:
        - path: /
          backend:
            serviceName: myapp-internal-svc
            servicePort: 8080

###################
# tls secret #####
###############

apiVersion: v1
kind: secert
metadata:
  name: myapp-tls-cert
  namespace: default
data:
  tls-crt: base64 encoded cert
  tls-key: base64 encided key
type: kubernates.io/tls     ### have to be this type

Helm – package manager for kubernates – package YAML files

Helm charts – bundle of YAML files

HELM is templating engine also

helm search <whatever-you-need>
helm install <chart-name>	tamplate files will be filled from value.yaml
helm install –values=my-values.yaml <chartname>	values injection into tamplate files

Volumes – – 3 components

• Persistent volume – – PV
• Persistent volume Claim – – PVC
• Storage Class – – SC

it can be NFS server, Local disk or cloud. They are NOT namespaced. They are accessible from all namespaces. DB always use remote storage.

#### NFS example ####

apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv-name
spec:
  capacity:
    storage: 20Gi
    volumeMode: Filesystem
    accessModes:
      - ReadWriteOnce
    persistenVolumeReclaimPolicy: Recycle
    storageClassName: slow
    mountOptions:
      - hard
      - nfsvers=4.0
    nfs:
      path: /dir/path/on/nfs/server
      server: nfs-server-ip-address

#### Google Cloud example ####
##############################

apiVersion: v1
kind: PersistentVolume
metadata:
  name: test-vol
  lebels:
    failure-domain.beta.kubernates.io/zone: us-centrall-a__us-centrall-b
spec:
  capacity:
    storage: 400Gi
  accessModes:
  - ReadWriteOnce:
  gcePersistentDisk:
    pdName: my-beta-disk
    fsType: ext4

##### Local Storage Example #####
#################################

apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv-example
spec:
  capacity:
    storage: 40Gi
  volumeMode: Filesystem
  accessModes:
  - ReadWriteOnce:
  persistenVolumeReclaimPolicy: Delete
  storageClassName: local-storage
  local:
    path: /mnt/disks/ssd1
  modeAffinity:
    required:
     nodeSelectorTerms:
     - matchExpressions:
       - key: kubernates.io/hostname
         operator: In
         values:
         - example-node

#################################
# PVC - Persistent Volume Claim # have to exist in same namespace
#################################

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: pvc-name          <-- here you refer to
spec:
  storageClassName: manual
  volumeMode: Filesystem
  accessModes:
    - ReadWriteOnce
  resources:
    request:
      storage: 40Gi
____________________________________
# you have to use pvc in pod config #
____________________________________

kind: pod
metadata:
  name: my-app
spec:
  containers:
    - name: my-app
      image: nginx
      volumeMounts:
      - mountPath: "/var/www/html"  (volume is monuted into container)
        name: mypd
  volumes: 
    - name: mypod
      persistantVolumeClaim:
        claimName: pvc-name  <-- this is reference

Storage Class provisions Volumes dynamically when PVC claims it

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: storage-class-name   <--- reference
provisioner: kuernates.io/aws-ebs
parameters:
  type: io1
  iopsPerGB: "10"
  fsType: ext4

#### you have to claim it via PVC ###

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: pvc-name         
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    request:
      storage: 40Gi
 storageClassName: storage-class-name  <--reference

Stateful Set

• Fix Individual DNS name for each pod
• $[pod name].$[governating service domain]
• IP can change, name stays same

Services — static IP for each pod

• ClusterIP – default – internal (node range) cubectl get pod -o wide
• Headless – communicete with 1 specific pod. Use case: Stateful app, like DB. Cluster IP: none. Use DNS
• NodePort – External traffic has access to fix port on each worker node. Instead of ingress (30000-32768 range) just for test. Not production.
• LoadBalancer – external access through cloud provider, extension of NodePort and ClusterIP

kubectl get endpoints ### same name as service ###
                      ### keeps track of which pods are members/endpoints of the service ####

DaemonSet – Caluculates how many replicas are needed based on existing nodes. 1 replica per node.

### use pipe for multiline strings and > for one line string###

multilineString: | 
  svaba tralala
  olala itd
  i tako sledece nedelje
  u isto vreme