CIVO-*: free 250 creditg for learning [https://civo.com/thedigitallife](https://www.youtube.com/redirect?event=video_description&redir_token=QUFFLUhqbDl6WXBET1dIcl9ZVFZlU1JoR3JBMXZwUVRld3xBQ3Jtc0ttYmJ2ZU9kcFpxbkRNYmllcGl1enVRNmM0UDMzdkZZSE45aDlrYmtUV2xLYTlvbjNVYUZmdXRBSkRrN19fWnNSQXJvS21CQ1hidGtZekRQSU5vZEpGQ2xkUnROcnAxQ3ZoUmZ5cFhNUjFvdG9mQ0JBQQ&q=https%3A%2F%2Fcivo.com%2Fthedigitallife&v=pumX2Ds5L0c)
single server is might be not enough to run all containers, we deploy all containers on multiple servers, or cluster.
To handle the traffic on all the server we need to use LOAD BALANCER and route to all our server instead of just one. we also should have some kindf of shared storage.
The Solution is KUBERNETES.
- EDCD: Distrubuted key-values pair, that store workerload, namespaces, different configurations. only accesible from api server of kubernetes
- API: every communication for kubernetes, recieve all the request and act as frontend to cluster. we can use termincal client
- controller manager: Run worker loads and process in the background and it regulates the shared state of the cluster and perform routines. e.g if you change any configuration then contoller mamanger spot changes and instructs other comonennt what to do.
- Schedular: controller manger instructs the schedular and schedular is reposnible for assigning the resources.
- kubelet: take instructions from api server and take care if the all the worker load are healthy and containers running in desired state and reports backs to master, its just taking instruction and executing them.
- kube-proxy: also runnning on the worker load and deals networking and exposing services to externals. the proxy is handling the trafic and making changes in the OS to reciebes the those reqs.
- Pod: smallest entity of K8s, atleast one container in it.
Commands
# set default namespace
k config set-context --current --namespace webserver
# get all pods
k get all pods -o wideNginx webserver:
# create deployment.yml file with following content
apiVersion: apps/v1
kind: Deployment
metadata:
name: webserver
namespace: webserver
labels:
app: webserver
spec:
selector:
matchLabels:
app: webserver
replicas: 2
template:
metadata:
labels:
app: webserver
spec:
containers:
- name: nginx
image: nginx:latest
ports:
- containerPort: 80
# use "kubectl apply -f deployment.yml" to apply this deployment
# now if you want to make available your service to public internet
# then you have to create service object
apiVersion: v1
kind: Service
metadata:
name: servicename
namespace: namespace
spec:
selector:
app: appname
# ---
# type: ClusterIP
# ClusterIP means this service can be accessed by any pod in the cluster
# ports:
# - name: http
# port: 8080
# targetPort: 80
# protocol: TCP # optional protocol
# ---
# type: NodePort
# NodePort means this service is only accessible by pods in the same namespace
# ports:
# - name: http
# port: 80
# nodePort: 30001
# protocol: TCP # optional protocol
# ---
type: LoadBalancer
# LoadBalancer means this service is load-balanced across all nodes in the cluster
ports:
- name: http
port: 80
targetPort: 30001
protocol: TCP # optional protocol
# after you apply service the digital ocean is automatically going
# to create load balancer for you and then just hit the external api.
ahsan@ahsan-XPS-13-7390:~/k-configurations/nginx-deployment$ **kubectl get service**
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
webserver LoadBalancer 10.245.102.12 206.189.249.236 80:31470/TCP 2m55s
Persistent Storage:
mounting specific folder inside the pod running from host drive
# yml file with persistent volumes:
apiVersion: apps/v1
kind: Deployment
metadata:
name: webserver
namespace: webserver
labels:
app: webserver
spec:
selector:
matchLabels:
app: webserver
replicas: 2
template:
metadata:
labels:
app: webserver
spec:
containers:
- name: nginx
image: nginx:latest
ports:
- containerPort: 80
volumeMounts:
- name: local
mountPath: /usr/share/nginx/html
volumes:
- name: local
hostPath:
path: /home/ahsan/k-configurations/nginx-deployment/nginx
above the shared storage that is not accrose the all pods: for this we need to have some shared storeage available to all pods. for that we can use goodle cloud storage aws s3 or azure but for now we are using nfs service for storage. create file with nfs-pv.yml .
**apiVersion: v1
kind: PersistentVolume
metadata:
name: nfs
spec:
capacity:
storage: 500Mi
accessModes:
- ReadWriteMany
storageClassName: nfs
nfs:
server: 192.168.1.7
path: "/srv/nfs"**kubectl apply -f nfs-pv.yml and use kubectl get pv to get persistent volumes
Config maps in kubernetes:
to store keyvalues secrets and the enviromental variables
apiVersion: v1
kind: ConfigMap
metadata:
name: webserver-cm
data:
# key: value
# file: |
# content ==> this will create file
# ---
nginx.conf: |
user nginx;
worker_processes 1;
events {
worker_connections 10240;
}
http {
server {
listen 80;
server_name _;
location /test {
return 200 'This is the content you want to return';
add_header Content-Type text/plain;
}
location / {
autoindex on;
root /usr/share/nginx/html;
index index.html index.htm;
}
}
}