# Cedacri CKAD Mock Exam - ALL

## Before we start

- Install autocomplete
- `sudo -c 'kubectl completion bash >/etc/bash_completion.d/kubectl'`
- Install aliases
- `echo 'alias k=kubectl' >>~/.bashrc`
- `echo 'alias kx=kubectx' >>~/.bashrc`
- `echo 'alias kbs=kubens' >>~/.bashrc`
- `echo 'complete -F __start_kubectl k' >>~/.bashrc`
- Install jq
- `sudo apt-get install -y jq`
- Install yq
- `sudo wget https://github.com/mikefarah/yq/releases/download/v4.13.4/yq_linux_amd64 -O /usr/bin/yq && sudo chmod +x /usr/bin/yq`

## Victims order

1. massimiliano ortenzi
1. stefano riccardi
1. gianmarco tonelli
1. angela ruscitto
1. stefano lombardi

## Question 0 - Namespaces

Create a namespace called `alpha-x123555`.

After the namespace creation, create a list of all the existing namespaces in `/home/workshop/namespaces.txt`.
The list should contain only the names of the namespaces, one per line:

## /home/workshop/namespaces.txt
alpha-x123555
default
everything-works
ingress-nginx
kube-node-lease
kube-public
kube-system
local-path-storage

### Solution

TBD

## Question 1 - Pod

Deploy a pod with name `nginx-xh78` with the image `registry.sighup.io/workshop/nginx:alpine` in the default namespace.
The container inside the pod should be named `nginx-xh78-container`.

After the pod creation, write a script `/home/workshop/get_status.sh` that uses `kubectl` to retrieve the status of the `nginx-xh78` pod when invoked.

### Solution

TBD

## Question 2 - Pod Labels

Deploy a pod with name `child` with the image `registry.sighup.io/workshop/redis:alpine` in the default namespace.
The pod shoud have a label `fruit=pineapple`.

After the pod creation, create a replicaset called `parent` that will adopt the pod `child`.

Check the desired and current pods in the replica set are correct.
Try to delete the pod `child` and see if another one is created.

### Solution

TBD

## Question 3 - Pod Placement

Deploy a pod with name `busybox-jke3` with the image `registry.sighup.io/workshop/busybox` in the default namespace.
The pod should be scheduled **only** on the **master nodes**.
The solution should work in case the number of master nodes increase in the future.

Do not edit the master node definition.

### Solution

TBD

## Question 4 - Static pod

Create a namespace `static`.

Create a static pod in the first worker node in the namespace `static`.
The pod should:

- be called `static-pod`
- run the image `registry.sighup.io/workshop/busybox`
- execute the command `sleep 1000`

### Solution

TBD

## Question 5 - Job

Create a namespace `red`.

Create a job called `red` in the `red` namespace.
The job should run the image `registry.sighup.io/workshop/busybox` and execute `sleep 2 && echo done`.
The job should run 10 times and execute at most 3 runs in parallel.

Check the jobs log when terminated.

### Solution

TBD

## Question 6 - CronJob

Create a `rooster` CronJob that everyday at `6:00 AM` executes `date; echo chicchirichi`.
You can use the `registry.sighup.io/workshop/busybox` image in the definition.

### Solution

TBD

## Question 7 - Expose deployment internally via ClusterIP

Create the namespace `beta`.

Create a deployment `cache` with the label `flavour=cache` that uses the image `registry.sighup.io/workshop/redis:alpine` in the namespace `beta`. The container inside the pod template definition should expose port `6379`.

Expose the deployment inside the cluster with a service `cache-service` on the port `6379` .

After the deployment and service are created, scale the number of replicas of the `cache` deployment to 3.

### Solution

TBD

## Question 8 - Expose deployment externally via NodePort

Create the namespace `hello`.

Create a deployment `hello-world` with the label `app=hello` that uses the image `gcr.io/google-samples/node-hello:1.0` in the namespace `hello`. The container inside the pod template definition should expose port `8080`.

Expose the deployment outside the cluster with an appropriate service `hello-service`.
The service should be mapped on the port `30003` of the nodes.

### Solution

TBD

## Question 9 - Deployment update and rollback

1. Create a deployment in the default namespace with the image `registry.sighup.io/workshop/nginx:1.7.9` with 3 replicas called `nginx`.

2. Execute a rolling update saving the change cause to the image `registry.sighup.io/workshop/nginx:1.9.9`.

3. In case of problems rollback to the previous version.

### Solution

TBD

## Question 10 - Set requests and limits

Create the namespace `blue`.

Create a deployment `blue` with `3` replicas in the `blue` namespace that uses the image `registry.sighup.io/workshop/httpd:latest`. The container should be named `blue-container` and have memory request of `20Mi` and a memory limit of `50Mi`.

### Solution

TBD

## Question 11 - Troubleshooting applications

Inside the namespace `everything-works` there is a `website` deployment which is currently not working.
Identify the problem and fix it.

### Solution

TBD

## Question 12 - InitContainer and Probes

1. Create a pod in the namespace `default` called `slowstart`. The pod should mount an `emptyDir` volume called `shared` at `/usr/share/nginx/html`. Moreover, the pod should have a container called `nginx` that:

- runs the image `registry.sighup.io/workshop/nginx`
- has a liveness probe that performs an `httpGet` on port `80` at `/filedinamico.html`

2. Run the pod, it should be in `CrashLoopBack` state as the liveness probe is failing.

3. Add an initContainer called `init` that mounts the `shared` volume and create the file `/usr/share/nginx/html/filedinamico.html`. You can use the image `registry.sighup.io/workshop/busybox`

### Solution

TBD

## Question 13 - Sidecar

Create a pod called `writer-reader` in the `default` namespace with following specifications:

- Use an emptyDir volume called `shared`.
- Have a container `writer` running `registry.sighup.io/workshop/busybox` that mounts the `shared` volume at `/opt/app_logs/` and writes `hello` in a file `/opt/app_logs/wave.log`
- Another container `reader` running `registry.sighup.io/workshop/ubuntu` that outputs the file created by the other container to stout.

Extract the first 10 lines of the logs of the `reader` container at `/home/workshop/multi.logs`.

### Solution

TBD

## Question 14 - DNS Resolution

Create a deployment in the namespace `default` called `apache` that uses the image `registry.sighup.io/workshop/httpd:latest`.
The container inside the pod template definition should expose port `80`.

Expose the deployment internally with a service `apache-service` on port `8080`.

Verify the DNS resolution of the `apache-service` via `nslookup` using a temporary pod running the image `registry.sighup.io/workshop/busybox`. Save `nslookup` output at `/home/workshop/dnsresolution.txt`

### Solution

TBD

## Question 15 - Statefulsets

Create a statefulset called `whoami` in the namespace `default` containing 3 pods running the image `registry.sighup.io/workshop/ubuntu`.

Each pod should mount the following persistentVolumeClaim at the mountPath `/pvc`:

- name: `whoami`
- accessModes: `"ReadWriteOnce"`
- storage: `1Gi`

The first pod of the three should output every 10 seconds to stdout: `"I am the leader"`
The other pods should output every 10 seconds to stdout: `"I am a follower"`

### Solution

TBD

## Question 17 - Network policy

0. Create the namespace `policy`.

1. Create a pod `busybox`:

- with image `registry.sighup.io/workshop/busybox:latest`
- with stag `type=client`
- in the namespace `policy`
- that executes the command `sleep 3600`

2. Create a pod `cannot`:

- with image `registry.sighup.io/workshop/busybox:latest`
- with stag `type=hacker`
- in the namespace `policy`
- that executes the command `sleep 3600`

3. Create a deployment `httpd`:

- with the image `registry.sighup.io/workshop/httpd:latest`
- in the namespace `policy`
- that exposes the `containerPort` 80

4. Expose the deployment internally `httpd` via a service at port `80`.

5. Create a network policy that allows the `busybox` to talk to the `httpd` deployment on the port `80` but blocks the access from all other pods.

### Solution

TBD

## Question 20 - Configmap

1. Create a ConfigMap called `beta-5000` in the namespace `default` with the following values:

- `COLOR=red`
- `FLAVOUR=garlic`

2. Create a pod `configmap-reader` in the namespace `default` that uses the ConfigMap `beta-5000`, mounting `COLOR` and `FLAVOUR` as environment variables inside a container that echos these values every minute.

You can use the `registry.sighup.io/workshop/busybox` image for the container inside the `configmap-reader` pod.

### Solution

TBD

## Question 21 - Secret

1. Create a secret called `secret-3fg` in the namespace `default` containing the following `secret.config` file:

Hello world
Doing Kubernetes stuff

Create a `secret-reader` pod in the namespace `default` that mounts this secret in `/opt/secret.config` and output its content every minute.

### Solution

TBD

## Question 22 - Secret Token of Service Account

Create a service account called `luke` in the namespace `default`.
Retrieve the service account token and write the base64 **decoded** token to file `/home/workshop/token`

### Solution

TBD

## Question 31 - Volumes

1. Create a persistent volume claim `alpha-claim` in the namespace `default` with:

- storageClass `local-path`
- access mode `ReadWriteOnce`
- Capacity `1Gi`

2. Create a pod called `volume-user` that uses the image `registry.sighup.io/workshop/nginx:alpine` that mounts this volume on `/usr/share/nginx/html`.

3. Enter the pod `volume-user` and create a file `index.html` inside the mounted directory with arbitrary content.

4. Delete and recreate the pod

5. Check that the file `/usr/share/nginx/html/index.html` inside the pod is still present.

### Solution

TBD

## Question 40 - jsonpath

Use the `-o jsonpath` to extract the `osImage` from all the nodes as save it in `/home/workshop/osImage_list.txt`

> TIP: `osImage` is inside `status->nodeInfo`.

### Solution

TBD