Kubernetes demo

This repo has the kubernetes concept demo for PODS, Deployments, StatefulSets, DaemonSets, services and Ingress

PODS

The pods.yml file contains the tasks for pods creations.

Task 1:

You should create a pod by following the requirements:

• Pod Name: nginx-pod
• Pod Image: nginx:alpine
• Pod Label: app=nginx
• Namespace: default
• Container Port: 80

Task 2:

Please find a pod called save-me and save its manifest (yaml output format) to $HOME/k8s_pods/save-me-pod.yml.

Task 3:

A new Pod web has been deployed in namespace: trouble. It has failed. Please find it, figure out the reason and fix it.

Requirements:

• Image: nginx:1.19-alpine
• Pod Status: Running

Task 4:

A new redis-db pod has been deployed in namespace trouble. It failed. Please fix it.

Task 5:

Create a new pod named redis with redis:123 image. And yes, the image name is wrong!

Requirements:

• Name: redis
• Image Name: redis:123
• Namespace: default Do not forget to copy secret phrase, test will fail after next task.

Task 6:

Now please change that redis pod image to the correct version.

Requirements:

• Name: redis
• Image Name: redis:5-alpine

Task 7:

Create a Pod (namespace: default, name: envtest, image: busybox:1.34, container process: env && sleep infinity). Specify the following ENV variables for the main container:

STUDENT_FIRST_NAME: <yourname> STUDENT_LAST_NAME: <yourlastname> Check Pod logs with kubectl logs command. Should be similar to this:

... 
STUDENT_FIRST_NAME=Ivan 
STUDENT_LAST_NAME=Ivanov 
...

Save actual POD’s logs (full output) into $HOME/k8s_pods/default-envtest.log file

Task 8 (advanced):

Create a Pod (namespace: default, name: i-know-who-i-am, image: busybox:1.34, container process: env && sleep infinity). Specify following ENV variables for the main container:

MY_NODE_NAME should contain name of pod node MY_POD_NAME should contain name of pod (do not provide it manually) MY_POD_NAMESPACE should contain name of pod namespace (do not provide it manually) MY_POD_IP should contain pod IP address MY_POD_SERVICE_ACCOUNT should contain a pod service account

Task 9:

Delete all pods in clean-up namespace.

Task 10(advanced)

Create a static pod.

Requirements:

Pod name: nginx-static Pod image: nginx:alpine Pod label: app=nginx-static Namespace: static Container Port: 80 How to verify: Try to delete static pod. It should be recreated

Do not forget to copy secret phrase, test will fail after next task.

Task 11(advanced):

Delete a static pod nginx-static

Deployments

Task 1:

Create a new deployment called nginx-deploy:

Requirements: Name: nginx-deploy Image: nginx:1.19-alpine Namespace: default Replicas: 1 Labels: app=nginx-deploy Do not forget to copy secret phrase, test will fail after task 3.

Task 2:

Inspect the details listed below, add necessary options to kubectl create deploy command to produce following deployment configuration:

Name: easy-peasy Image: busybox:1.34 Replicas: 5 Command: sleep infinity

Task 3:

Scale nginx-deploy deployment to 6 replicas.

Task 4:

Inspect the details listed below and create deployment:

Deployment Name: -app Replicas: 1 Deployment Labels: task: deploy app: -app student: Pod(s) Labels: deploy: -app kind: redis role: master tier: db Container: Image: redis:5-alpine Port: 6379 Name: redis-master Init Container: Image: busybox:1.34 Command: sleep 10

Task 5:

Current deployment nginx-deploy release has nginx:1.19-alpine image.New release should use nginx:1.21-alpine.Use rolling update process

Task 6:

A lemon deployment has been deployed in namespace trouble, but there are no pods associated with it. Figure out the root cause and fix the issue.

Task 7:

A orange deployment has been deployed in namespace trouble, but it doesn’t work properly. Figure out the root cause and fix the issue.

StatefulSets

Task 1:

Create a new StatefulSet:

Requirements:

• Name: random-generator
• Image: sbeliakou/random-generator:1
• Namespace: default
• Replicas: 3
• Service: random-generator
• Labels: app=random-generator Do not forget to copy secret phrase, test will fail after task 3.

Task 2:

Add volumeClaimTemplates to random-generator sts. Recreate statefulset if it’s needed.

• Name: logs
• mountPath: /logs Capacity: 10Mi accessMode: ReadWriteOnce

Task 3:

Update container’s image to version 2. Please pay attention to the way how StatefulSet recreates pods. It starts from 2 and goes to 0.

Task 4:

Run any test pod. Using nslookup check the record of random-generator service. Save the output of the below commands to $HOME/k8s_sts.txt

nslookup random-generator-0.random-generator.default.svc.cluster.local
nslookup random-generator-1.random-generator.default.svc.cluster.local
nslookup random-generator-2.random-generator.default.svc.cluster.local

DaemonSets

Task 1:

Create a new node and set up a worker role for it.

Remember to copy secret phrase, test will fail after task 5.

Task 2:

Taint your nodes.

Requirements:

• control-plane:
  • key: node-role.kubernetes.io/control-plane
  • effect: NoSchedule
• worker:
  • key: node-role.kubernetes.io/worker
  • effect: NoSchedule Do not forget to copy secret phrase, test will fail after task 6.

Task 3:

Deploy daemonset

Requirements:

• Name: fluentd-elasticsearch
• Image: quay.io/fluentd_elasticsearch/fluentd:v4 Please pay attention how many pods were created and why.

Do not forget to copy secret phrase, test will fail after next task.

Task 4:

Modify fluentd-elasticsearch DaemonSet to ensure that pods run on every node.You should change only DaemonSet configuration!

Do not forget to copy secret phrase, test will fail after next task.

Task 5:

Create one more node and set up worker role for it.

Please pay attention on number of pods.

Do not forget to copy secret phrase, test will fail after task 6.

Task 6:

jq utility should be installed before running checker!

Get details of nodes in JSON format and save it to $HOME/nodes-info.json file. Delete all worker nodes. Untaint control-plane node. Remove fluentd-elasticsearch DaemonSet.

Services

Task 1:

Create pod-info-svc service for pod-info-app deployment:

Requirements:

• Name: pod-info-svc
• Type: ClusterIP
• Service Port: 80
• Service TargetPort: 80 Investigate pod-info-app deployment and choose selector and do not change the deployment.

Task 2:

Run a pod (based on image busybox:1.28) and execute the following commands (as given below) inside this pod and save outputs into files:

wget -q -O- pod-info-svc  
$HOME/testing-clusterip-web.log
nslookup pod-info-svc
$HOME/testing-clusterip-nslookup.log

Task 3:

Create deployment myapp, image: sbeliakou/web-pod-info:v1, replicas: 1 Create headless service myapp-headless pointing to myapp pods Create non-headless service myapp-clusterip pointing to myapp pods Checking name resolution:

Non-headless service has own IP address (and port), and proxies traffic to backends:

$ kubectl run --rm -it test --image=busybox:1.27 --restart=Never nslookup myapp-clusterip

Headless service doesn’t proxy traffic to backends, it simply responds (on early dns lookup phase) with the IP address(es) where to go:

$ kubectl run --rm -it test --image=busybox:1.27 --restart=Never nslookup myapp-headless

Task 4:

We have already created hello-hello web application for you. Create a new service to access this web application, check the requirements. Figure out all necessary settings from the deployment.

Requirements:

• Name: hello-hello-service
• Type: NodePort
• Downstream Pod Port (Service targetPort): 80
• Node Port: 30300 For verification you can execute curl $NODE_IP:30300 command or open the browser page http://$NODE_IP:30300. Substitute $NODE_IP with the IP address of your node.