Potato2

Senior Design Final Project, Kubernetes on EC2 instances.

Tools/technology being used

• AWS
  • CLI
  • EC2
• Docker
• Git
• Kubernetes
  • Kubectl
  • Kops
  • Minikube
• Virtualbox/VMware

Diagram of infrastructure

How to execute the code

Setting up Kubernetes on a Virtual Machine

Install VirtualBox or VMWare

1. Run the following command:

   sudo apt-get install virtualbox
   

2. Make sure that Intel VT-x/EPT or AMD-V/RVI is enabled.

Install Kubectl

1. Run the following commands:

   sudo apt-get update && sudo apt-get install -y apt-transport-https curl
   curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -
   echo "deb https://apt.kubernetes.io/ kubernetes-xenial main" | sudo tee -a /etc/apt/sources.list.d/kubernetes.list
   sudo apt-get update
   sudo apt-get install -y kubectl
   

Install Docker

1. Run the following command:

   sudo apt-get install docker
   

Install Minikube

1. Run the following commands:

   curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
   sudo install minikube-linux-amd64 /usr/local/bin/minikube
   

2. Make sure Intel VT-x/EPT or AMD-V/RVI is enabled.

Run Minikube

0. Run the following commands:

   sudo chown -R $(id -u):$(id -g) ~/.minikube/*
   
   sudo chown -R $(id -u):$(id -g) ~/.kube/*
   

1. Run this command to start Minikube

   minikube start
   

2. Run these commands to check that the cluster is running:

   kubectl get nodes
   kubectl cluster-info
   

3. Start the minikube dashbaord:

   minikube dashboard
   

4. Visit your minikube dashboard: http://127.0.0.1:34191/api/v1/namespaces/kube-system/services/http:kubernetes-dashboard:/proxy/#!/cluster?namespace=default

Setting up Kubernetes on a AWS using Windows 10

Installing the requirements

1. Download and install AWS CLI: https://aws.amazon.com/cli/
2. Download and move kubectl: https://kubernetes.io/docs/tasks/tools/install-kubectl/ then move it to the directory C:\Kube
3. Download kops: https://github.com/kubernetes/kops, then rename the download to kops.exe. Finally move the file to C:\Kube
4. Add AWS CLI, kubectl, and kops to path:
   1. Press the windows key then type in env
   2. Select the option that says Edit the system environment variables
   3. Click on Environment Variables
   4. Under System Variables, double click Path.
   5. Select new and add C:\Program Files\Amazon\AWSCLI\bin
   6. Select new and add C:\Kube
5. Restart your computer.

Setting up the Kubernetes Cluster

1. Run the following command in order to create a S3 bucket for kops to use.

   aws s3api create-bucket --bucket potato2-kops-state-store --region us-east-1
   

2. Enable versioning on the newly created bucket:

   aws s3api put-bucket-versioning --bucket potato2-kops-state-store  --versioning-configuration Status=Enabled
   

3. Set environment varaibles for kops to use:

   SET KOPS_CLUSTER_NAME=potato2.k8s.local
   SET KOPS_STATE_STORE=s3://potato2-kops-state-store
   

   Note: in order to print out these variables in your CLI, run the following command: echo %KOPS_CLUSTER_NAME% echo %KOPS_STATE_STORE%

4. Create a public key for kops

   ssh-keygen
   
   kops create secret --name potato2.k8s.local sshpublickey admin -i C:\Users\yourusername/.ssh/id_rsa.pub
   

   Note: run ssh-keygen and select all the default options. It should save your private and public keys to C:\Users\yourusername/.ssh/id_rsa

5. Generate the cluster configuration. Creates the configuration and writes to the s3 bucket:

   kops create cluster --node-count=2 --node-size=t2.medium --zones=us-east-1a 
   

   Note: a lot of information will be displayed after this step.

6. Build the cluster:

   kops update cluster --name potato2.k8s.local --yes
   

7. Before running this step, WAIT a couple minutes or you WILL get errors. It takes time to launch the nodes. You can check the progress of the build by logging into your AWS Console then going to EC2 and making sure you're in the same region that you listed in step 5.

   kops validate cluster
   

8. Check that the nodes are running:

   kubectl get nodes
   

Setting up the Kubernetes Dashboard

1. Run the following to start the dashboard

   kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/v1.10.1/src/deploy/recommended/kubernetes-dashboard.yaml
   

2. Get the AWS Hostname:

   kubectl cluster-info
   

3. Visit your dashboard using the command above or use the following command to create a proxy:

   kubectl proxy
   

   Then visit the url: http://localhost:8001/api/v1/namespaces/kube-system/services/https:kubernetes-dashboard:/proxy/

   Note: if you use the AWS Hostname the username will be admin and the password can be found with the following command:

   kops get secrets admin --type secret -oplaintext
   

4. When logging into the dashboard using the proxy method, it will ask you to enter a token. Obtain this token using this command:

   kops get secrets admin --type secret -oplaintext
   

Creating a service and/or deployment file using yaml

1. Create an nginx deployment file called run-my-nginx.yaml and put in the following:

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: nginx-deployment
     labels:
       app: nginx
   spec:
       replicas: 2
       selector:
         matchLabels:
           app: nginx
       template:
         metadata:
           labels:
             app: nginx
       spec:
         containers:
         - name: nginx
           image: nginx:1.7.9
           ports:
           - containerPort: 80
   

2. Create a mysql persistent volume file called mysql-pv.yaml and put in the following:

   kind: PersistentVolume
   apiVersion: v1
   metadata:
     name: mysql-pv-volume
     labels:
       type: local
   spec:
     storageClassName: manual
     capacity:
       storage: 20Gi
     accessModes:
       - ReadWriteOnce
     hostPath:
       path: "/mnt/data"
   ---
   apiVersion: v1
   kind: PersistentVolumeClaim
   metadata:
     name: mysql-pv-claim
   spec:
     storageClassName: manual
     accessModes:
       - ReadWriteOnce
     resources:
       requests:
         storage: 20Gi
   

3. Create a mysql deployment file called mysql-deployment.yaml and put in the following:

   apiVersion: v1
   kind: Service
   metadata:
     name: mysql
   spec:
     ports:
     - port: 3306
     selector:
       app: mysql
     clusterIP: None
   ---
   apiVersion: apps/v1 # for versions before 1.9.0 use apps/v1beta2
   kind: Deployment
   metadata:
     name: mysql
   spec:
     selector:
       matchLabels:
         app: mysql
       strategy:
         type: Recreate
       template:
         metadata:
           labels:
             app: mysql
         spec:
           containers:
           - image: mysql:5.6
             name: mysql
             env:
               # Use secret in real usage
           - name: MYSQL_ROOT_PASSWORD
             value: password
           ports:
           - containerPort: 3306
             name: mysql
           volumeMounts:
           - name: mysql-persistent-storage
             mountPath: /var/lib/mysql
           volumes:
           - name: mysql-persistent-storage
             persistentVolumeClaim:
               claimName: mysql-pv-claim
   

Using the dashboard to deploy an application and/or service

1. Once you've logged onto the dashboard, click the button at the top right that says "+ Create"
2. Enter the yaml presented in the above section then click OK.

Destroy a cluster

1. Run the following command to destroy the cluster

   kops delete cluster --name potato2.k8s.local --yes
   

FAQ - Potential or known issues related to the setup or execution of code

Important Note:

A lot of issues can be avoided just by having your host machine be a *nix operating system. In order words: try not to use Windows if at all possible.

Question #1: minikube start command states "Unable to start VM"

• These are the causes of this issue:

  • You might be trying to start minikube in a local VM:
    1. This is due to virtualbox not being installed on the machine.
    2. VT-X/AMD-v is not enabled to allow nested VMs.

  Answer

  • This is fixed by doing the following:
    1. Install virtualbox by running the following command:

    sudo apt-get install virtualbox
    

    2. If you are running the VM in vmware then click on VM > Settings > Processors > Check Virtualize Intel VT-x/EPT or AMD-V/RVI then restart the VM.

Question #2: kubectl get nodes command states "The connection to the server localhost:8080 was refused"

• These are the causes of this issue:

  1. This happens because minikube has not finished creating the VM.

  Answer

  • This is fixed by doing the following:
    1. Just wait until minikube has finished.
    2. In the meantime check that the kubectl config is present and correct:
       1. Run the following command to both check for and display the contents of the kube config file:

       sudo cat ~/.kube/config
       

Question #3: minikube gets stuck on "Waiting for pods: apiserver"

• These are the causes of this issue:

  1. This happens for because minikube has not finished creating the VM.

  Answer

  • This is fixed by doing the following:
    1. Just wait until minikube has finished.

Question #4: minikube dashboard states "failed to open browser: exit status 3"

• These are the causes of this issue:

  1. This happens because you might have run the following command with sudo:

     sudo minikube dashboard
     

  Answer

  • This is fixed by doing the following:
    1. Don't run the command using sudo.

Question #5: minikube dashboard states "Error getting machine status: load: filestore: open ~/.minikube/machines/minikube/config.json: permission denied"

• These are the causes of this issue:

  1. This happens because you don't have permission to use that file.

  Answer

  • This is fixed by doing the following:
    1. Change directory into the .minikube directory then chown the entire directory.

    sudo chown -R $(id -u):$(id -g) *
    

    2. Change directory into the .kube directory then chown the entire directory.

    sudo chown -R $(id -u):$(id -g) *
    

Question #6: minikube dashboard states "kube-system:kubernetes-dashboard is not running: Temporary Error: Error getting service kubernetes-dashboard"

• These are the causes of this issue:

  1. This happens when minikube was not configured properly due to a permission issue. Make sure you follow step 0 in the documentation.

  Answer

  • This is fixed by doing the following:
    1. Run the following command:

    minikube delete
    

    then continue following the documentation from the beginning.

Question #7: running kops create cluster states "SSH public key must be specified when running with AWS"

• These are the causes of this issue:

  1. This happens because kops requires a SSH public key.

  Answer

  • This is fixed by doing the following:
    1. Create the public key:

       kops create secret --name potato2.k8s.local sshpublickey admin -i C:\Users\yourusername/.ssh/id_rsa.pub
       

Question #8: running kops create secret states "error adding SSH public key: error fingerprinting SSH public key"

• These are the causes of this issue:

  1. This happens because you need to have a .ssh folder.

  Answer

  • This is fixed by doing the following:
    1. Run the following command to create the .ssh folder with a id_rsa public and private key:

       ssh-keygen
       

    2. Use all the default options and take note of where the key was saved to. Usually:

       C:\Users\yourusername/.ssh/id_rsa.pub
       

Question #9 running kops get secret sates "State Store: Required value: Please set the --state flag or export KOPS_STATE_STORE."

• These are the causes of this issue:

  1. This happens because the value of KOPS_STATE_STORE was not properly set.

  Answer

  • This is fixed by doing the following:
    1. Run the following command:

       SET KOPS_STATE_STORE=s3://potato2-kops-state-store
       

Question #10 Logging into the kubernetes dashboard gives me errors saying "configmaps is forbidden: User cannot list..."

• These are the causes of this issue:

  1. This happens because the user you logged in as does not have the permissions to view the metrics listed.

  Answer

  • This is fixed by doing the following:
    1. make sure you are listing the admin key: ie. the username is after secrets keyword below

    kops get secrets admin --type secret -oplaintext
    

Question #11 How do I access the website on the cluster?

Answer

• This is done by doing the following:

  1. Log onto the dashboard and click on Services then click on the service you want to access.
  2. Click on the ip address under External endpoints.

  OR

  3. Run the following command:

     kubectl get svc
     

  Note: No matter which option you take, you'll have to make sure there's an external ip.

Question #12 Why do I not have an external IP on my deployment/service?

• These are the causes of this issue:

  1. This happens because you haven't exposed the port/ip.

  Answer

  • This is fixed by doing the following:
    1. Run the following command:

       kubectl expose deployment nginx-deployment --port 80 --type=LoadBalancer
       

    2. List all the services with the following command:

       kubectl get services -o wide
       

Question #13 I keep getting "server IP address could not be found" when trying to access the website hosted by the cluster.

• These are the causes of this issue:

  1. This happens when you are running the wrong command or looking at the wrong area on the dashboard.

  Answer

  • This is fixed by doing the following:
    1. See Questions 11 and 12 for the answer.