Shulou(Shulou.com)06/03 Report--

This article mainly explains "how developers understand kubernetes". The content of the explanation is simple and clear, and it is easy to learn and understand. Please follow the editor's train of thought to study and learn "how developers understand kubernetes".

Overview

Using the docker run command in JAVA development with a self-built Docker repository makes it easy to deploy JAVA services, but there are several problems with deploying applications using Docker:

A docker run is not a reliable way to deploy a service because the containers it creates run on a single machine. Although the Docker engine provides some basic administrative functions, such as automatically restarting the container when the container crashes or the calculator restarts. But it can't handle machine crashes. The high availability of the service cannot be guaranteed! Another problem is that services usually do not exist in isolation, but are interdependent, such as databases and message queues. We usually need to deploy or undeploy the service and its dependencies as a unit.

A particularly useful method in the development process is to use Docker Compose. Docker Compose is a tool that allows you to define a set of containers declaratively using YAML files, and then start and stop them as groups.

But the obvious problem with using Docker Compose is that it is limited to one machine. To deploy services reliably, you must use a Docker orchestration framework, such as Kubernetes.

Introduction to Kubernetes

Kubernates is a Docker orchestration framework and a software layer on top of Docker that transforms a set of computer hardware resources into a single pool of resources for running services. It strives to maintain the number of instances required for each service and to ensure that they are always online, even if the service instance or machine crashes. Combining the flexibility of containers with the complexity of Kubernetes is a powerful way to deploy services.

Kubernetes has three main functions:

Resource management: treat a group of computers as a resource pool of CPU, memory, and storage volumes, and treat a computer cluster as a single computer. Scheduling: select the machine on which to run the container. By default, scheduling takes into account the resource requirements of the container and the resources available for each node. It can also implement the deployment of affinity containers on the same node, or keep specific containers dispersed on different nodes (anti-affinity, anti-affinity) service management: implement the concept of naming and versioning services, which can be directly mapped to specific services in the micro-service architecture. The orchestration framework ensures that the required number of normal instances is always run. It implements load balancing of requests. The orchestration framework can also perform rolling upgrades to the service and allow you to roll back to the old version. Kubernetes architecture Kubernetes architecture

Kubernetes runs on a set of machines. The computer roles in the Kubernetes cluster are divided into master nodes and ordinary nodes. There are only a few primary nodes (perhaps only one) and many ordinary nodes in the cluster.

The master node is responsible for managing the cluster. The "normal node" of the Kubernetes is called the "worker node" and it runs one or more Pod. Pod is the deployment unit of Kubernetes and consists of a set of containers.

The primary node runs multiple components, including the following:

API server: the REST API used to deploy and manage services, for example, can be used by the kubectl command line. Etcd: the NoSQL database that stores cluster data key values. Scheduler: select the node on which you want to run POD. Controller manager: run the controller to ensure that the cluster state matches the expected state. For example, a controller called a replication controller ensures that the required service instance is running by starting and terminating the instance.

A normal node runs multiple components, including the following:

Kubelet: create and manage the Pod running on the node. Kube-proxy: manages the network, including load balancing across Pod. Pods: application service.

Next, let's take a look at the key Kubernetes concepts that need to be mastered to deploy services on Kubernetes. Mastering these concepts captures the core of Kubernetes.

Key Concepts of Kubernetes

Kubernetes is complex, but once you have mastered some concepts of "key objects", you can use Kubernetes efficiently. Kubernetes defines many types of objects, and from the developer's point of view, the most important objects are as follows:

Pod:

Pod is the basic deployment unit of Kubernetes. It consists of one or more containers that share IP addresses and storage volumes. The pod of a service instance usually consists of a single container, such as a container running JVM. In some cases, however, the Pod contains one or more "sidecar" containers that implement supporting functions. For example, a Nginx server can have a side car container that periodically executes git pull to download the latest version of the site. The life cycle of Pod is short because the container of Pod or the node on which it is running may crash.

Deployment:

Deployment: the declarative specification of Pod. Deployment is a controller that ensures that the required number of Pod instances (service instances) is always running. It supports version control by rolling upgrades and rollbacks.

Service:

A static / stable network address provided to the client of the application service. It is a form of service discovery provided by the infrastructure. Each Service has an IP address and a DNS name that can be resolved to that IP address, and load balancing TCP and UDP traffic across one or more Pod. The IP address and DNS name can only be accessed within Kubernetes.

Service defaults to ClusterIp mode, and if you need external access to this Service, you need to use two other types of objects: NodePort and LoadBalancer.

ConfigMap:

A named collection of name and value pairs that defines the externalized configuration of one or more application services. The definition of a Pod container can refer to ConfigMap to define the environment variables of the container. It can also use ConfigMap to create configuration files within the container. You can use Secret to store sensitive information, such as passwords, which is also a form of ConfigMap.

Thank you for your reading, the above is the content of "how developers understand kubernetes". After the study of this article, I believe you have a deeper understanding of how developers understand kubernetes, and the specific use needs to be verified in practice. Here is, the editor will push for you more related knowledge points of the article, welcome to follow!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.