Kubernetes - Deploy Guestbook example on Kubernetes
This scenario explains how to launch a simple, multi-tier web application using Kubernetes and Docker. The Guestbook example application stores notes from guests in Redis via JavaScript API calls. Redis contains a master (for storage), and a replicated set of redis 'slaves'.
Core Concepts
The following core concepts will be covered during this scenario. These are the foundations of understanding Kubernetes.
Pods
Replication Controllers
Services
NodePorts
Step 1 - Start Kubernetes
To start we need a running Kubernetes cluster. The details of this are explained in the Launch Kubernetes cluster scenario.
Start a single-node cluster using the helper script. The helper script will launch the API, Master, a Proxy and DNS discovery. The Web App uses DNS Discovery to find the Redis slave to store data.
launch.sh
Health Check
Check everything is up using the following health Check:
kubectl cluster-info kubectl get nodes
If the node returns NotReady then it is still waiting. Wait a couple of seconds before retrying.
Step 2 - Redis Master Controller
The first stage of launching the application is to start the Redis Master. A Kubernetes service deployment has, at least, two parts. A replication controller and a service.
The replication controller defines how many instances should be running, the Docker Image to use, and a name to identify the service. Additional options can be utilized for configuration and discovery. Use the editor above to view the YAML definition.
If Redis were to go down, the replication controller would restart it on an active node.
Create Replication Controller
In this example, the YAML defines a redis server called redis-master using the official redis running port 6379.
The kubectl create command takes a YAML definition and instructs the master to start the controller.
kubectl create -f redis-master-controller.yaml
What's running?
The above command created a Replication Controller. The Replication
kubectl get rc
All containers described as Pods. A pod is a collection of containers that makes up a particular application, for example Redis. You can view this using kubectl
kubectl get pods
Step 3 - Redis Master Service
The second part is a service. A Kubernetes service is a named load balancer that proxies traffic to one or more containers. The proxy works even if the containers are on different nodes.
Services proxy communicate within the cluster and rarely expose ports to an outside interface.
When you launch a service it looks like you cannot connect using curl or netcat unless you start it as part of Kubernetes. The recommended approach is to have a LoadBalancer service to handle external communications.
Create Service
The YAML defines the name of the replication controller, redis-master, and the ports which should be proxied.
kubectl create -f redis-master-service.yaml
List & Describe Services
kubectl get services kubectl describe services redis-master
Step 4 - Replication Slave Pods
In this example we'll be running Redis Slaves which will have replicated data from the master. More details of Redis replication can be found at http://redis.io/topics/replication
As previously described, the controller defines how the service runs. In this example we need to determine how the service discovers the other pods. The YAML represents the GET_HOSTS_FROM property as DNS. You can change it to use Environment variables in the yaml but this introduces creation-order dependencies as the service needs to be running for the environment variable to be defined.
Start Redis Slave Controller
In this case, we'll be launching two instances of the pod using the image kubernetes/redis-slave:v2. It will link to redis-mastervia DNS.
kubectl create -f redis-slave-controller.yaml
List Replication Controllers
kubectl get rc
Step 5 - Redis Slave Service
As before we need to make our slaves accessible to incoming requests. This is done by starting a service which knows how to communicate with redis-slave.
Because we have two replicated pods the service will also provide load balancing between the two nodes.
Start Redis Slave Service
kubectl create -f redis-slave-service.yaml kubectl get services
Step 6 - Frontend Replicated Pods
With the data services started we can now deploy the web application. The pattern of deploying a web application is the same as the pods we've deployed before.
Launch Frontend
The YAML defines a service called frontend that uses the image _gcr.io/googlesamples/gb-frontend:v3. The replication controller will ensure that three pods will always exist.
kubectl create -f frontend-controller.yaml
List controllers and pods
kubectl get rc kubectl get pods
PHP Code
The PHP code uses HTTP and JSON to communicate with Redis. When setting a value requests go to redis-master while read data comes from the redis-slave nodes.
Step 7 - Guestbook Frontend Service
To make the frontend accessible we need to start a service to configure the proxy.
Start Proxy
The YAML defines the service as a NodePort. NodePort allows you to set well-known ports that are shared across your entire cluster. This is like -p 80:80 in Docker.
In this case, we define our web app is running on port 80 but we'll expose the service on 30080.
kubectl create -f frontend-service.yaml kubectl get services
Step 8 - Access Guestbook Frontend
With all controllers and services defined Kubernetes will start launching them as Pods. A pod can have different states depending on what's happening. For example, if the Docker Image is still being downloaded then the Pod will have a pendingstate as it's not able to launch. Once ready the status will change to running.
View Pods Status
You can view the status using the following command:
kubectl get pods
Find NodePort
If you didn't assign a well-known NodePort then Kubernetes will assign an available port randomly. You can find the assigned NodePort using kubectl.
kubectl describe service frontend | grep NodePort
View UI
Once the Pod is in running state you will be able to view the UI via port 30080. Use the URL to view the pagehttps://2886795360-30080-ollie02.environments.katacoda.com
Under the covers the PHP service is discovering the Redis instances via DNS. You now have a working multi-tier application deployed on Kubernetes.
We'll discuss NodePort in future scenarios.
Kubernetes - Deploy Guestbook example on Kubernetes的更多相关文章
- centos7下kubernetes(5。部署kubernetes dashboard)
基于WEB的dashboard,用户可以用kubernetes dashboard部署容器话的应用,监控应用的状态,执行故障排查任务以及管理kubernetes各种资源. 在kubernetes da ...
- 【Kubernetes学习之二】Kubernetes集群安装
环境 centos 7 Kubernetes有三种安装方式:yum.二进制.kubeadm,这里演示kubeadm. 一.准备工作1.软件版本 软件 版本 kubernetes v1.15.3 Cen ...
- Kubernetes 实战 —— 02. 开始使用 Kubernetes 和 Docker
创建.运行及共享容器镜像 P23 运行容器 P24 运行 P24 可以运行 Docker 客户端可执行文件来执行各种 Docker 命令.例如:可以试着从 Docker Hub 的公共镜像仓库拉取.运 ...
- kubernetes进阶(02)kubernetes的node
一.Node概念 Node是Pod真正运行的主机,可以物理机,也可以是虚拟机. 为了管理Pod,每个Node节点上至少要运行container runtime(比如docker或者rkt). kube ...
- kubernetes入门(04)kubernetes的核心概念(1)
一.ReplicationController/ReplicaSet 在Kubernetes集群中,ReplicationController能够确保在任意时刻,指定数量的Pod副本正在运行.如果Po ...
- 高可用Kubernetes集群-15. 部署Kubernetes集群统一日志管理
参考文档: Github:https://github.com/kubernetes/kubernetes/tree/master/cluster/addons/fluentd-elasticsear ...
- Kubernetes - Deploy Containers Using YAML
In this scenario, you'll learn how to use Kubectl to create and launch Deployments, Replication Cont ...
- kubernetes入门(05)kubernetes的核心概念(2)
一.使用 kubectl run 创建 pod(容器) 命令 kubectl run类似于 docker run,可以方便的创建一个容器(实际上创建的是一个由deployment来管理的Pod): 等 ...
- 高可用Kubernetes集群-14. 部署Kubernetes集群性能监控平台
参考文档: Github介绍:https://github.com/kubernetes/heapster Github yaml文件: https://github.com/kubernetes/h ...
随机推荐
- 亚马逊首次推出卖家APP 可掌握商品盈利状况
美国零售巨头亚马逊近日首次对外发布了第一款针对卖家和商户的客户端,帮助他们更加高效的管理商品和销售数据. 据美国科技新闻网站 Mashable 报道,之前亚马逊在商户移动客户端方面一直空缺,许多商户不 ...
- [转]oracle数据库定时任务dbms_job的用法详解
这篇文章给大家详细介绍了dbms_job的用法,用于安排和管理作业队列,通过使用作业,可以使ORACLE数据库定期执行特定的任务.有需要的朋友们可以参考借鉴. 一.dbms_job涉及到的知识点 ...
- mysql常用语句入门整理
这篇属于小白入门级别,如果你已经高手可以直接跳过 1.运行数据库mysqld.exe,客户端直接mysql -uroot(root是默认用户名) -p 2 showdatabases,showtabl ...
- oracle selinux 问题
Error: cannot restore segment prot after reloc: Permission Denied http://www.oracledistilled.com/ora ...
- SSH新学,关于面向对象的看法
流程:model-->dao-->service-->impService-->action 如果只是操作单个的一个表,比如user表,则都写到user的流程中 如果要操作俩个 ...
- asp.net如何实现跟踪检查用户知否查看了邮件。
有时我们有这样一种需求场景,我们给很多用户发了邮件,需要一个反馈,用户是否查看了我们发送的邮件,百度了以下果然有方案. 我总结实践了下这个过程,同时有自己的一点使用感受.记录下希望对你有帮助. 有人想 ...
- 蜗牛慢慢爬 LeetCode 2. Add Two Numbers [Difficulty: Medium]
题目 You are given two non-empty linked lists representing two non-negative integers. The digits are s ...
- win8平板APP开发的教程文章
http://blog.csdn.net/tcjiaan/article/details/7866595 基于C#的Metro工程如何引用C++的动态库——FIleNotFound解决办法: http ...
- JSON和Django内置序列化
JSON 什么是JSON JSON 指的是 JavaScript 对象表示法(JavaScript Object Notation) JSON 是轻量级的文本数据交换格式 JSON 独立于语言 * J ...
- UVA11736_Debugging RAM
题目绝对够水,我就不详细说明了. 直接上代码吧.只是提示一下要用 unsigned long long. (不知道我不用字典树为什么会超时,肿么搞的) #include <iostream> ...