Kubernetes has become an essential part of industries and is changing the world of technology. Janet Kuo, the co-chair of KubeCon, addressed the Kubernetes features and its importance in the coming time. She talks about the Kubernetes extensions and solutions that will change the world.
Top Machine Learning Courses & AI Courses Online
This cheat sheet will provide a reference to working professionals in the implementation of Kubernetes. It contains all the necessary details to help beginners who are new to Kubernetes. The Kubernetes Cheat Sheet is a quick and handy medium to refer to all the essential concepts and commands of Kubernetes.
Enrol for the Machine Learning Course from the World’s top Universities. Earn Masters, Executive PGP, or Advanced Certificate Programs to fast-track your career.
An open-source platform for automatic deployment and scaling containers across the clusters of hosts to provide container-centric infrastructure is known as Kubernetes (also known as “Kube” or k8s). It allows easy and efficient management of different hosts running Linux containers by clustering them.
Kubernetes is a platform that is designed for managing the life cycle of containerized applications and services completely. A Kubernetes user can define the ways in which an application should run and interact with different applications.
Trending Machine Learning Skills
Users can switch traffic between different versions of applications, perform updates, scale up and down the services, etc. with Kubernetes. It offers users a high degree of flexibility, reliability, and power in managing applications.
Some of the major features of Kubernetes are:
- Maximize resources by making better use of hardware.
- A container orchestrator across multiple hosts.
- Automate the deployment process and updates.
- Able to run a Linux container.
- Auto-scaling helps in launching containers on cluster nodes.
- Scaling up and down as required.
- Self-healing by replacing, rescheduling, and restarting the dead containers.
- Automated rollbacks and rollouts.
- Load balancing and service discovery.
- Auto-restart, auto-placement, and auto-replication, etc.
Read: Deep Learning Algorithm
Architecture of Kubernetes
The architecture of Kubernetes consists of layers: Higher and lower layers. The complexity of abstracting the higher layer can be found in the lower layers. The individual physical or virtual machines are brought together into a cluster. A shared network is used for communication between each server. So, just like other distributed platforms, Kubernetes has one master (at least), and multiple compute nodes.
- The master of Kubernetes schedules the deployments, exposes the API and manages the overall Kubernetes cluster.
- The node runs a container runtime, agent for communicating with the master, and other monitoring components, logging, etc.
Components of Kubernetes architecture
Let’s have a look at the purpose and components of master and nodes in the Kubernetes architecture.
The master maintains the desired state of the cluster. Since it manages the whole cluster, it is called master. It contains:
- API server: Kubernetes API server
- Scheduler: Used for pod scheduling in worker nodes
- Controller: Manages pod replication
- Etcd: A metadata service
It contains necessary services that are important for running the pods. The master manages the nodes. It is also called Minion. It contains:
- Pod: Group of containers
- Docker: Container-based technology, user space of OS.
- Kubelet: Container agents that are responsible for maintaining the set of pods.
- Kube-proxy: Routes traffic coming into a node from the service
Now, let’s understand the important commands of Kubernetes.
Kubectl is the command-line tool for Kubernetes. The basic Kubectl commands can be divided into:
- Pod and Container Introspection
- Cluster Introspection
- Quick Commands
Pods and Container Introspection
|For describing pod names||Kubectl describe pod<name>|
|For listing all current pods||Kubectl get pods|
|For listing all replication controllers||Kubectl get rc|
|For showing the replication controller name||Kubectl describe rc <name>|
|For listing replication controllers in a namespace||Kubectl get rc –namespace=”namespace”|
|For showing a service name||Kubectl describe svc<name>|
|For listing services||Kubectl get cvc|
|For watching nodes continuously.||Kubectl get nodes -w|
|For deleting a pod||Kubectl delete pod<name>|
|For getting version-related information||Kubectl version|
|For getting configuration details||Kubectl config g view|
|For getting cluster-related information||Kubectl cluster-info|
|For getting information about a node||Kubectl describe node<node>|
|For displaying metrics for a pod||Kubectl top pod|
|For displaying metrics for a node||Kubectl top node|
|For watching Kubelet logs||Watch -n 2 cat/var/log/kublet.log|
|For getting logs from the service for the container||Kubectl logs -f<name>>[-c< $container>]|
|For the execution of the command on service by selecting a container||Kubectl exec<service><commands>[-c< $container>]|
The below quick commands are often used and hence, very useful.
|For launching a pod with a name and an image.||Kubectl run<name> — image=<image-name>|
|For creating a service described in <manifest.yaml>||Kubectl create -f <manifest.yaml>|
|For scaling the replication counter to count the number of instances.||Kubectl scale –replicas=<count>rc<name>|
|For mapping the external port to the internal replication port.||Expose rc<name> –port=<external>–target-port=<internal>|
|For stopping all pods in <n>||Kubectl drain<n>– delete-local-data–force–ignore-daemonset|
|For creating a namespace.||Kubectl create namespace <namespace>|
|For allowing the master node to run pods.||Kubectltaintnodes –all-node-role.kuernetes.io/master-|
Some of the familiar objects used in Kubernetes are as follows:
|List of Common Objects|
|cm= conf gmaps||Cluster role bindings|
|ev= events||hpa= horizontal pod autoscaling|
|jobs||ds= daemon sets|
|No = nodes||ns= namespaces|
|po= pods||Pod preset|
|Psp= pod security policies||Pv= persistent volumes|
|quota= resource quotas||rs= replica sets|
|roles||rc= replication controllers|
|sc= storage classes||pdb= pod distribution budgets|
|crd=custom resource definition||Pod templates|
|csr= certificate signing requests||sa= service accounts|
|Netpol- network policies||Role bindings|
|ing= ingress||pvc= persistent volume claims|
|ep=end points||sts= stateful sets|
Also Read: Regularization in Deep Learning
Kubernetes Command Cheat Sheet
All the basic information about Kubernetes, it’s architecture and commands are shown in below Kubernetes cheat sheet:
Popular AI and ML Blogs & Free Courses
If you’re interested to learn more about deep learning techniques, machine learning, check out IIIT-B & upGrad’s PG Diploma in Machine Learning & AI which is designed for working professionals and offers 450+ hours of rigorous training, 30+ case studies & assignments, IIIT-B Alumni status, 5+ practical hands-on capstone projects & job assistance with top firms.
What is Kubernetes?
Kubernetes is an open-source platform for automating deployment, scaling, and management of containers. A container is a ready-to-run software package that contains everything required to run the application: code, runtime environment, application, and system libraries. The containers are like logical units of application, and when Kubernetes groups them, it becomes easy to manage and discover the containers. These containers are deployed and scaled across hosts to provide a container-centric infrastructure. In short, it is a platform that manages the entire lifecycle of containerized applications.
What are the main components of Kubernetes architecture?
There are two main components in the Kubernetes architecture- master node and worker node. A node is the main worker machine which is important for running a group of clusters called pods. The whole cluster is managed by the master node, whereas the worker node performs the tasks assigned by the master node. These nodes have many inbuilt components within them. The master node contains Kubernetes API Server, Kubernetes Scheduler, Kubernetes Controller, and Etcd. The worker node has pod, docker, kubelet, and kube-proxy.
What do kubelet and kube-proxy do?
Kubelet, which is present in every node, ensures that containers are running well in a pod. It does so by taking a set of PodSpecs provided mainly through apiserver and checking the containers present in those PodSpecs. A PodSpec is a YAML or JSON object that describes a pod. Kubernetes Network Proxy, also known as kube-proxy, maintains network rules on nodes. Communication from network sessions inside or outside the cluster to pods is possible through these network rules. Kube-proxy forwards the traffic if the OS packet filtering layer is not available.