Search the Community

A free and open-source system called Kubernetes manages the container solutions like Docker. With the help of Docker, you may build the application containers from predefined images. The next stage is provided by Kubernetes which enables you to operate on numerous containers across several platforms and equalize the load among them. You may follow the instructions in this manual to set up Kubernetes on Ubuntu 20.04 via two different methods. Update the Ubuntu Packages Before installing Kubectl on Ubuntu, we need to update all the packages of the Ubuntu system. For an update, we have to utilize the “Apt” utility of Ubuntu within the “update” command as follows: $ sudo apt update Set Up Docker The first prerequisite to properly install Kubernetes on Ubuntu is to set up Docker. To install Docker, we utilize the “apt” utility of Ubuntu within the installation command. The Docker package should be mentioned as “docker.io” followed by the “-y” option to force the “apt” utility to install Docker. The Docker will be installed in a while, and you can start using it after enabling it. $ sudo apt install docker.io -y To enable Docker on the Ubuntu system, we utilize the “systemctl” utility. The two separate commands are utilized with the “enable” and “status” keywords to enable the Docker service and check its status after enabling it. The output of the status query displays that the Docker service is running perfectly fine. $ sudo systemctl enable docker $ sudo systemctl status docker After enabling the docker utility, you have to start using the same “systemctl” utility followed by the “Start” keyword and service name as “docker”. $ sudo systemctl start docker Install Kubernetes Before installaing Kubernetes, it’s necessary to install the “curl” utility in your system so that you can add the key of Kubernetes in your system. The “Snap” package is utilized to install the “curl” utility. The utility will be installed successfully. $ sudo snap install curl After the installation of the “curl” utility, we utilize it within the command followed by the “fsSL” option to get the GPG key of Kubernetes from its official cloud repository. This command gets the signing key as depicted in the output image: $ curl -fsSL https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo tee /usr/share/keyrings/kubernetes.gpg Now, to add the Kubernetes repository within the default repositories of our system, we have to use the “echo” command followed by the “deb” instruction. Ensure that your other Ubuntu node has the Docker installed and you have to execute the following instruction on other nodes, i.e., for smooth communication: $ echo “deb [arch=amd64 signed-by=/usr/share/keyrings/Kubernetes.gpg] http://apt.kubernetes.io/ kubernetes-xenial main” | sudo tee -a /etc/apt/sources.list A utility called Kubeadm, or Kubernetes Administrator, aids with cluster initialization. Utilizing the community-sourced standards expedites setup. The work bundle that launches the containers on each node is called Kubelet. You may access the clusters via the command line with the program. Run the following instructions on every server node to install the Kubernetes installation tools via the “snap” utility for each instruction followed by the “—classic” keyword: $ sudo snap install kubelet --classic $ sudo snap install kubectl --classic $ sudo snap install kubeadm --classic After the successful installation of Kubernetes tools, we utilize the version instructions for each tool to look for the installed version as follows: $ kubeadm version $ kubelet version $ kubectl version Configure Kubernetes It’s time to configure the installed tools of Kubernetes on our system. In this section, you may learn how to get the servers ready for a Kubernetes deployment. Perform the following commands on each Linux machine that you used as a node. First of all, turn off the swap storage first. Run the “swapoff” command with the “-a” option to carry out this operation. The “swapoff” command should be followed by the “sed” command as follows: $ sudo swapoff -a $ sudo sed -I ‘/ swap / s/^\(.*\)$/#\1/g’ /etc/fstab It’s time to load the modules of “containerd”. To load those modules, we open the “containerd” configuration file using the GNU nano editor. This file is in the “etc” folder per the following command: $ sudo nano /etc/modules-load.d/containerd.conf Upon opening the configuration file, we load the “overlay” and “br_netfilter” modules by adding the following two keywords: overlay br_netfilter After mentioning the modules in the files, we have to utilize the “modprobe” instruction followed by the names of modules to load them finally. $ sudo modprobe overlay $ sudo modprobe br_netfilter It’s high time to configure the network of Kubernetes by utilizing its configuration file located in the “etc” folder. The GNU nano editor is utilized to open the file. $ sudo nano /etc/sysctl.d/kubernetes.conf We need to set the following shown variables for Kubernetes to enable the networking. This configuration sets up the iptables for Kubernetes. Now, make sure to save the file before exiting. net.bridge-bridge-nf-call-ip6tables = 1 net.bridge-bridge-nf-call-iptables = 1 net.ipv4.ip_forward = 1 Now, we reload the system services and configurations once and for all. The “sysctl” utility is utilized here with the “—system” option. You will see that the new configurations are added and reloaded, i.e. displayed in the output. $ sudo sysctl --system After successfully loading the Kubernetes modules and reloading the services, we have to assign a unique name to each node in your Kubernetes network. For instance, we want to set the current node as the master node of the Kubernetes network. Therefore, we try the “hostnamectl” utility in the command to set the hostname as “master-node”. $ sudo hostnamectl set-hostname master-node The worker node is also set using the same instruction with a different node name; this command needs to be performed at the worker server. $ sudo hostnamectl set-hostname omar Open the host configuration files on every node within our Kubernetes network and add the IP addresses of each node within the file. You need to mention the host names along with IP addresses to identify them uniquely. $ sudo nano /etc/hosts The image depicts the IP addresses and their hostnames in the host configuration file that is opened via the nano editor. Make sure that you are at the master node right now and open the Kubelet configuration file via the nano editor. $ sudo nano /etc/default/kubelet Make sure to write the following shown line into the file to set a KUBELET_EXTRA_ARGS variable for Kubernetes on the master node. KUBELET_EXTRA_ARGS=”—cgroup-driver=cgroupfs” Now, you have to reload the just-set configuration within your master node and the worker nodes. The “systemctl” utility is utilized with the “daemon-reload” keyword. After providing the password, don’t forget to restart the Kubernetes service. $ systemctl daemon-reload Conclusion In the end, your Kubernetes service is successfully installed and configured. Make sure to reload the Docker service and perform the specified commands on the worker nodes as well. View the full article

In software development, building, managing, deploying, and shipping the application in containers is one of the most used and isolated approaches. Different container runtimes are used to containerize the applications such as Docker engine, CRI-O, Podman, containerd, and so on. To manage the containers in container runtime, roll out changes, monitor the applications, or operate multiple container runtimes, the developers usually consider and use Kubernetes. In this article, we will demonstrate: What is Kubernetes? Kubernetes Architecture How to Get Started With Kubernetes and Kubectl Using Docker Desktop? How to Get Started With Kubernetes and Kubectl Using Minikube? Conclusion What is Kubernetes? Kubernetes is a free open-source container composition or orchestration (coordination of multiple containers) system that is used to develop, manage, and deploy applications and software. The Kubernetes is used by the developers to automate the operational task, roll back the changes, and monitor and scale the applications and software. Kubernetes Architecture Kubernetes offers an adaptable and flexible architecture that consists of different components. It is used to discover the services and maintain loose connections within the Kubernetes cluster. The major components of Kubernetes architecture are listed below: Kubernetes Cluster Kubernetes Control Plane Kubernetes Node Kubernetes Pod Kubernetes Cluster When the user starts working with the Kubernetes, it creates a cluster known as the Kubernetes cluster. It is a primary component of Kubernetes. It consists of a Master and multiple slave node or at least one node that acts as master and slave at one time. The master node is also known as the Control Plane in Kubernetes used to manage the slave node and schedule the container efficiently. In contrast, the slave node manages the containers in pods, Kubelets, and Kube-proxy: Kubernetes Control Plane The Kubernetes Control Plane is a master node in the Kubernetes cluster that is responsible for making decisions for the cluster such as scheduling the containers or nodes according to configuration, exposing API, and communicating with working nodes: The Kubernetes control plane contains the following components: kube scheduler kube controller manager cloud controller manager kube-api-server etcd kube scheduler The kube scheduler is one of the control plane components that is used to monitor the newly created pod that currently is not part of any node. It selects or allocates the node to that pod to execute it. It is responsible for scheduling the resource requirements, applying hardware, software, or policy constraints, data locality, deadlines, and inter-workload interference. kube controller manager The kube controller manager is used to run the controller processes. Actually, it runs the different processes or controllers within a single binary process. Its main task is to match or align the desired and current state of the object within the node. There are different types of controllers and some of them are listed below: Node Controller: It is used to manage the working nodes. Basically, it notices the node and responds when the node goes down. Job Controller: It is used to monitor the idle job or task that needs to be run. Then, it creates a pod to fulfill the task. Replication Controller: It monitors and maintains the desired container count within the replication cluster. Service Control Controller: It creates and manages the new service account for the new namespace. EndpointSlice Controller: It is used to populate the endpoint object to provide the link between pods and services. cloud controller manager The cloud controller manager is another important element of the Kubernetes control plane that is only available in the cluster that is running on the user cloud provider. If users are using the Kubernetes on a personal computer, the cluster does not include any cloud controller manager. The cloud controller manager is used to embed the cloud-specific control logic and enable the user to link the cluster to their cloud provider’s API. The cloud controller manager also has different types of controllers such as “Node Controller” to manage the node, “Route Controller” for setting up the route, and “Service Controller” to manage the services of the cloud provider such as creating, deleting or updating load balancer. kube-api-server The Kube api server is considered as a front end of the Control plane through which all components within a cluster can interact. It sends the REST operations, performs administrative tasks, and exposes the Kubernetes API. etcd The etcd is considered as the storage unit of the Control plane. It is a distributed, consistent Key value store that is used to store all cluster data such as cluster state and configurations. It is the place where Kubernetes stores all its information. Kubernetes Node Kubernetes nodes are also known as working nodes that provide the Kubernetes runtime environment. These nodes are responsible for managing and running applications within container runtimes. These nodes are controlled and managed by the Control plane(master node). Each node can operate and run multiple pods and containers: The major components of the Kubernetes node are listed below: Kube proxy Kubelet Pod Kube proxy The kube proxy in Kubernetes is a network proxy that executes on each node of the Kubernetes cluster. It manages and controls the network rules and allows the network communication to Pod or containers from network sessions within or outside of the Kubernetes cluster. Kubelet Kubelet is an agent of Kubernetes that runs on each node of the Kubernetes cluster. It is basically used for interaction purposes and is also responsible for the execution of containers within pods. It makes sure that containers described in pods are executed and in a healthy state. The communication between the Control plane and the working nodes is also done through kubelet. Pod Pods are another major component of the Kubernetes cluster that execute multiple container applications or software. Kubernetes Pod Kubernetes Pod is another small deployable component of the Kubernetes cluster that executes within the working node. Each Kubernetes node can execute multiple pods. Each Kubernetes Pod consists of multiple containers with shared network and storage resources. These pods provide isolation just like the container’s isolation. Within each Kubernetes pod, the two major components exist that are: Volume: Volume is referred to as a shared file system or storage option that persists the container data. It is an external storage mechanism that connects the containers to a persistent storage unit. Container: Containers are lightweight executable packages that encapsulate that program along with dependencies and configurations. It is an isolated virtual environment that uses OS level of virtualization and shares OS among applications: How to Get Started With Kubernetes and Kubectl Using Docker Desktop? To get started with Kubernetes, users need to start the Kubernetes node. For that particular purpose, users can use the Docker Desktop application as it is one of the well-liked container runtimes. To start development with Kubernetes, first, install the required tools and start playing with Kubernetes with the help of the following steps: Step 1: Install Docker Desktop Step 2: Install Kubernetes Step 3: Install kubectl Step 4: Use kubectl to Start and Run First Pod Step 1: Install Docker Desktop Docker Desktop is the GUI platform and containerization system that permits developers to build, run, and ship applications and software in an isolated environment. Users can use it as a Kubernetes container runtime as well to start the Kubernetes node inside the Kubernetes cluster. To install and set up a Docker environment on Windows, go through our associated “Install Docker Desktop on Windows” article. Step 2: Install Kubernetes After installing the Docker Desktop application, launch the app and install Kubernetes by enabling it from the Docker Desktop settings. For illustration, follow the provided procedure. Launch Docker Desktop Application First, launch the Docker Desktop application by searching it in the Windows Startup menu: Navigate to Docker Settings Next, click on the “Settings Gear” icon to open the Docker Desktop settings: Install Kubernetes Navigate to the “Kubernetes” option from the left pane. After that, mark the below pointed “Enable Kubernetes” icon and hit the “Apply & restart” button: Upon doing so, the “Kubernetes Cluster Installation” box will appear on the screen. Hit the “Install” button to install Kubernetes: The below output indicates that we have successfully installed and started the Kubernetes cluster using the Docker Desktop application: Step 3: Install Kubectl The kubectl is a Kubernetes command line tool that is used to manage and operate the Kubernetes cluster. To perform any operation within Kubernetes, users are required to have the kubectl tool on their system. To install kubectl on Windows, first, download the “kubectl.exe” file. After that, set the Path environment variable to use the “kubectl” command in the Windows terminal. For illustration, go through the following steps. Download Kubectl For Windows To download the latest and stable version of the kubectl tool for Windows, navigate to the Kubernetes official website and click on the below-highlighted link: Alternatively, the user can use the “curl” command to download the “kubectl.exe” file. For this purpose, first launch the Windows terminal such as PowerShell or Command Prompt. After that, execute the below-given command: curl.exe -LO "https://dl.k8s.io/release/v1.28.3/bin/windows/amd64/kubectl.exe" Move Kubectl Execution File to C Drive By default, the tool will be downloaded in the “Downloads” directory. It is better to move its executable file to the “C” local disk. For this purpose, first, create a new folder named “kubectl” in the “C” local disk. After that, move the “kubectl.exe” file into the newly created directory: Open Environment Variable Control Panel Settings Next, launch the “Environment variable” Control panel settings by searching it in the Startup menu: Add Kubectl to Windows Path Variable In order to access the kubectl command from the Windows command line, users are required to add the kubectl path to an environment variable. For this purpose, click on the “Environment Variables” button from the “Advanced” menu: Click on the “Path” option under the “System variables” and hit the “Edit” button: Next, first press the “New” button, paste the path where the “kubectl.exe” file exists, and hit the “OK” button: Verification For verification, launch the Windows PowerShell via the Start menu: After that, run the “kubectl” command to confirm if we have effectively installed the Kubectl or not: kubectl The below output indicates that we have installed the command successfully: Use kubectl Command to Check Current Context The “kubectl config current-context” command is used to access the current Kubernetes cluster context. When the user installs the Kubernetes on the system, this command will point to the Kubernetes cluster executing on the system. To access the current Kubernetes context, execute the below command. kubectl config current-context Here, you can see we are using Docker desktop to operate and run the Kubernetes server: Step 4: Use Kubectl to Create and Run the First Pod To run the first application in Kubernetes pod and to use the kubectl command, follow the below demonstration. Get Kubernetes Nodes First, access the current executing nodes using the “kubectl get nodes” command: kubectl get nodes The output shows that currently, only one node is running which is “docker-desktop” and it is a master node and also works as a working node: Create a New Pod Next, build a new pod to containerize the simple Nginx application. For this purpose, utilize the given command: kubectl run nginx --image=nginx --restart=Never In the above command, the “–image” option is used to specify the image that isolates the application within the container, and this container will execute inside the “nginx” pod. The “–restart” option is utilized to specify the restart policy of the Kubernetes pod: Get Pods After creating the kubectl pod, verify that if the pod is successfully running or not. For this purpose, access the Kubernetes pods using the “kubectl get pods” command: kubectl get pods Describe Pod In order to access the detailed configuration of the executing pod, describe the pod using the “kubectl describe pod <pod-name>” command: kubectl describe pod nginx Access Application Running in Pod To verify if the application is running inside the pod, the user needs to have the IP address of the pod where the application is executing. For this purpose, access the pod with complete status using the “-o wide” that shows the entire information of the pod including the IP address: kubectl get pods -o wide Access the Kubernetes Pod To access the application that is running inside the container of the pod, first access or login to the pod using the “kubectl exec -it <pod-name> bash” command: kubectl exec -it nginx bash Note: In the above output, this command is running but deprecated as well in kubectl utility and may be removed in an upcoming version. To run the latest command to access the Kubernetes pod, use “kubectl exec -it <pod-name> — <command>” command: kubectl exec -it nginx -- bash After accessing the pod shell, ping the IP address of the command to verify whether the Nginx application is effectively running inside the container or not: curl 10.1.0.6 The below output indicates that the application is effectively running inside the Kubernetes pod: That is all about getting started with Kubernetes and kubectl using the Docker Desktop application on Windows. How to Get Started With Kubernetes and Kubectl Using Minikube? To start working with Kubernetes and to use the kubectl command line utility to operate the Kubernetes cluster, users need to have a platform on which they can operate and execute the Kubernetes cluster. For this purpose, they can use the minikube Kubernetes implementation software. To get started with the Kubernetes cluster using minikube, follow the below-listed steps: Step 1: Install Minikube Step 2: Install kubectl Step 3: Enable Windows HyperV Step 4: Start Kubernetes Cluster or Node Step 5: Use Kubectl to Create and Run the First Pod Step 1: Install Minikube Minikube is a lightweight Kubernetes implementation that creates a simple virtual machine to execute the Kubernetes cluster that can only operate and run a single Kubernetes node. It is mostly used by developers for experimenting or testing purposes. It is also used by beginners to understand the Kubernetes works. To install the “minikube” on Windows, the user can use the following approaches: Method 1: Install Minikube Using Installer Method 2: Install Minikube Using PowerShell Method 1: Install Minikube Using Installer To install the minikube installer on Windows, first, go to minikube official website. Scroll down to the installation step, choose the “Windows” operating system, system architecture according to system specification, “Stable” release, and “.exe download” installer type. After that, click on the below highlighted “latest release” link to install the minikube installer: Run the Installer By default, the installer will be downloaded in the “Downloads” directory. After downloading, navigate to the “Downloads” directory, double-click on the minikube installer to execute it: Next, choose the language in which you want to access the installation configuration and instructions. Then, press the “OK” button: Install Minikube The “Kubernetes Minikube setup” window will appear on the screen. Hit “Next” to proceed with the installation process: First, accept the license agreement by clicking on the “I Agree” button: Next, browse the location where you want to install minikube and hit the “Install” button: After completing the installation process, hit the “Finish” button: This will install the minikube on the system and also automatically set its path to the Windows path variable. Verification To verify if we have installed the minikube on the system correctly or not, first launch the Windows PowerShell terminal as an administrator: Next, execute the “minikube” command. This should show you the minikube supported commands: minikube The below output shows that we have effectively installed the minikube on the system: Method 2: Install Minikube Using PowerShell To install minikube on Windows using the command line, first, launch the Windows PowerShell with administrative privileges. After that, execute the below commands. Create “minikube” Directory First, create a new directory in the Windows “C” disk. To do so, utilize the below command: New-Item -Path 'c:\' -Name 'minikube' -ItemType Directory -Force Download minikube Executable File Next, run the below command to download and install minikube on the system. In the below command, the new “minikube.exe” file will be created, and the configurations are written in that file from provided setup: Invoke-WebRequest -OutFile 'c:\minikube\minikube.exe' -Uri 'https://github.com/kubernetes/minikube/releases/latest/download/minikube-windows-amd64.exe' -UseBasicParsing Add minikube to Windows Path After installing the minikube, the user needs to add it to the Windows Path environment variable. For this purpose, execute the following commands: $oldPath = [Environment]::GetEnvironmentVariable('Path', [EnvironmentVariableTarget]::Machine) if ($oldPath.Split(';') -inotcontains 'C:\minikube'){ [Environment]::SetEnvironmentVariable('Path', $('{0};C:\minikube' -f $oldPath), [EnvironmentVariableTarget]::Machine) } Note: After completing the process, the user needs to close the Windows PowerShell and relaunch it to use minikube commands: For verification, execute the “minikube” command: minikube That is all about the installation of Minikube on Windows. Step 2: Install kubectl The kubectl is a Kubernetes command line tool that manages and operates the Kubernetes cluster. To install kubectl on Windows, follow the procedure that is described in the above section “How to Get Started With Kubernetes and Kubectl Using Docker Desktop?”. Step 3: Enable Windows HyperV Virtual support is necessary to run the Kubernetes cluster on minikube Windows. Because Minikube required the container or virtual machine to run the Kubernetes cluster on Windows. To use the Windows built-in virtualization (HyperV), users are required to enable it first on the system. To do so, go through the following instructions: Launch “Turn Windows features on or off” Settings First, open the Windows Start menu, search for the “Turn Windows features on or off” Control panel setting, and launch it from the search the results: Enable HyperV Virtualization Next, mark the below highlighted “Hyper-V” checkbox and hit the “OK” button: After enabling the Hyper-V on Windows, restart the system to implement the changes. Enable Virtualization Via PowerShell Alternatively, the user can enable the Hyper-V virtualization via Windows PowerShell. For this purpose, first, launch the PowerShell as administrator. After that, run the following command: Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V -All Step 4: Start Kubernetes Cluster or Node After completing the installation process of Minikube and kubectl, start the first instance or Kubernetes node inside the Kubernetes cluster. For this purpose, simply execute the “minikube start” command: minikube start At that point user may encounter some error and fail to start the Kubernetes instance on minikube as highlighted below: Resolve the “Failed to start hyperv VM” Error To resolve the “Failed to start hyperv VM” error, the user needs to delete the Kubernetes node from minikube using the “minikube delete” command: minikube delete After that, again run the “minikube start” command to start the new Kubernetes cluster: minikube start Here, you can see we have resolved the encountered error and effectively start the minikube cluster (Kubernetes cluster): Note: Minikube can run only one node in the Kubernetes cluster. The corresponding node will act as both the Control plane node and the working node. Verification For verification, check the Kubernetes node to verify if the Kubernetes cluster is executing on minikube or not: kubectl get nodes Here, you can see we have successfully started the “minikube” Kubernetes node on Windows: Step 5: Use Kubectl to Create and Run the First Pod Execute the first application in the Kubernetes Pod using kubectl command line utility by following the below instructions. Create a New Pod To create a new pod and to containerize the Nginx application, execute the below command: kubectl run nginx --image=nginx View the Pod IP Address For verification, view the Kubernetes pods using the “kubectl get pods” command. Note the IP address of the pod to access the encapsulated Nginx application: kubectl get pods -o wide Login to Minikube Cluster To access the application running in a pod, the user needs to access the Kubernetes node in which the pod is executing. To do so, use the “minikube ssh” command: minikube ssh Verify Application is Executing in Pod After accessing the shell of the Kubernetes node, utilize the “curl” command along with the IP address to verify if the Nginx default page is executing or not: curl 10.244.0.3 The output shows that we have successfully executed the new application within Pod of Kubernetes node: That is all about Kubernetes and kubectl on Windows. Conclusion To use Kubernetes and kubectl on Windows, users can run the Kubernetes cluster as a server on the Docker Desktop application and also on minikube. To get started with the Kubernetes cluster, users need to install Docker Desktop or minikube on the system. On Docker Desktop, enable and install the Kubernetes. On minikube, there is no need to install Kubernetes. Next, install the “kubectl” command line tool to operate the Kubernetes cluster and start playing with the Kubernetes cluster. This blog has illustrated how to get started with Kubernetes and kubectl on Windows. View the full article