用docker弹性部署自己的服务
很久不看docker的东西了，之前了解的一些基本命令都忘得差不多了，适逢工作需要，再来复习巩固下。今天想完成的是：借助docker不部署下自己的服务。
环境准备
都说“巧妇难为无米之炊”，所以还是需要先准备下的。
OS：Ubuntu 16.04， 2G内存
docker：1.13.2
coding language: golang (gin web framework)
编码
将服务跑起来，是我们要完成的第一个步骤，而且是最重要的一个步骤。所以这一步需要仔细调试，为了方便，我就写的简单点。
- app.go
package main
import (
"os"
"log"
"github.com/gin-gonic/gin"
"net/http"
"time"
)
func GetHostName() string {
hostname, err := os.Hostname()
if err != nil {
log.Fatal(err)
}
return hostname
}
func GetCurrentTime() string {
timer := time.Now()
return timer.String()
}
func startGinApp() {
app := gin.Default()
app.GET("/ping", func(context *gin.Context) {
context.JSON(http.StatusOK, gin.H{
"message": "当前为您服务的主机为：" + GetHostName(),
})
})
app.GET("/", func(context *gin.Context) {
context.JSON(http.StatusOK, gin.H{
"message": "当前时间为：" + GetCurrentTime(),
})
})
app.Run(":8080")
}
func main() {
startGinApp()
}
- 把服务跑起来
go run app.go
- curl一下看看服务是否正确跑起来了
➜ gin curl http://localhost:8080
{"message":"当前时间为：2018-10-14 11:31:23.016121853 +0800 CST m=+0.254631109"}%
➜ gin curl http://localhost:8080/ping
{"message":"当前为您服务的主机为：BiaodeMacBook-Pro.local"}%
➜ gin
制作Makefile
经过刚才的测试，代码可以正确跑起来了。但是要做到“一次编码，到处运行”，还是需要在构建阶段下点心思的，为了更好的维护，借助Makefile来规范构建过程，是比较合适的方法。
- Makefile
# 这个是注释
# 开头可以声明一大堆变量名
BUILD_NAME ?= httpserver
COMPILER ?= go
BUILD ?= build
# 上方留一个空格，区分变量区和构建区
all: build test deploy clean
build:
CGO_ENABLED=0 GOOS=linux GOARCH=amd64 $(COMPILER) $(BUILD) -o $(BUILD_NAME)
test:
echo "test over."
deploy:
echo "deploy over."
clean:
echo "clean over."
.PHONY: build test deploy clean
- 构建服务
➜ gin ls
Makefile app.go
➜ gin make all
go build -o httpserver
echo "test over."
test over.
echo "deploy over."
deploy over.
echo "clean over."
clean over.
➜ gin ls
Makefile app.go httpserver
➜ gin ./httpserver &
[1] 6450
➜ gin [GIN-debug] [WARNING] Creating an Engine instance with the Logger and Recovery middleware already attached.
[GIN-debug] [WARNING] Running in "debug" mode. Switch to "release" mode in production.
- using env: export GIN_MODE=release
- using code: gin.SetMode(gin.ReleaseMode)
[GIN-debug] GET /ping --> main.startGinApp.func1 (3 handlers)
[GIN-debug] GET / --> main.startGinApp.func2 (3 handlers)
[GIN-debug] Listening and serving HTTP on :8080
➜ gin curl http://localhost:8080/ping
[GIN] 2018/10/14 - 11:50:23 | 200 | 214.161µs | ::1 | GET /ping
{"message":"当前为您服务的主机为：BiaodeMacBook-Pro.local"}%
➜ gin
好了基本没什么问题，然后就可以通过scp命令将文件拷贝到linux服务器上了。
制作Dockerfile
由于golang构建出来的是二进制可执行程序，所以制作Dockerfile很简单。
- Dockerfile
FROM ubuntu:latest
MAINTAINER guopu [email protected]
WORKDIR /app
EXPOSE 8080
ADD . /app
CMD ["./httpserver"]
- 构建自己的镜像
root@Server218 /h/d/g/d/httpserver# docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
ginserver latest 6277a0180f4f 19 hours ago 101 MB
flaskindocker latest 170c6c0a41db 22 hours ago 131 MB
python 2.7-slim 14dad3ead5f4 4 days ago 120 MB
docker/compose 1.23.0-rc2 dc59a0b5e981 5 days ago 45.6 MB
ubuntu latest cd6d8154f1e1 5 weeks ago 84.1 MB
root@Server218 /h/d/g/d/httpserver# docker build -t httpserver .
Sending build context to Docker daemon 17.07 MB
Step 1/6 : FROM ubuntu:latest
---> cd6d8154f1e1
Step 2/6 : MAINTAINER guopu [email protected]
---> Running in 7106748df3ad
---> 0ae808029537
Removing intermediate container 7106748df3ad
Step 3/6 : WORKDIR /app
---> 7278bf9659e7
Removing intermediate container f7fdc76b19a8
Step 4/6 : EXPOSE 8080
---> Running in bedfabcb4b16
---> edf4c123f72f
Removing intermediate container bedfabcb4b16
Step 5/6 : ADD . /app
---> 36390e554a2f
Removing intermediate container b14cce9da53e
Step 6/6 : CMD ./httpserver
---> Running in 6682c8364717
---> b490fef8a9ca
Removing intermediate container 6682c8364717
Successfully built b490fef8a9ca
root@Server218 /h/d/g/d/httpserver# docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
httpserver latest b490fef8a9ca 4 seconds ago 101 MB
ginserver latest 6277a0180f4f 19 hours ago 101 MB
flaskindocker latest 170c6c0a41db 22 hours ago 131 MB
python 2.7-slim 14dad3ead5f4 4 days ago 120 MB
docker/compose 1.23.0-rc2 dc59a0b5e981 5 days ago 45.6 MB
ubuntu latest cd6d8154f1e1 5 weeks ago 84.1 MB
root@Server218 /h/d/g/d/httpserver#
- 让服务在docker中跑起来
root@Server218 /h/d/g/d/httpserver# docker run -d -p 8000:8080 httpserver
ebb1926206cdaf16eae9f4e13d5a7e70dd695da91463b438f77e45c6f65d3323
root@Server218 /h/d/g/d/httpserver# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
ebb1926206cd httpserver "./httpserver" 4 seconds ago Up 3 seconds 0.0.0.0:8000->8080/tcp nostalgic_wright
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：ebb1926206cd"}~
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/
{"message":"当前时间为：2018-10-14 04:14:29.116207751 +0000 UTC"}~
root@Server218 /h/d/g/d/httpserver#
至此，看起来服务已经成功在服务器环境下的docker中运行了。
弹性服务
一个容器跑一个服务，有些服务访问峰差很大的场景，就需要做下弹性适配，于是需要用一下docker swarm服务。这个在linux环境下需要进行安装。具体可以参考官网链接： https://github.com/docker/compose/releases
其运行以来一个YAML配置文件，具体细节不多说，上手吧。
- docker-compose.yml
version: "3"
services:
web:
image: httpserver:latest
deploy:
replicas: 5
resources:
limits:
cpus: "0.1"
memory: 50M
restart_policy:
condition: on-faliure
ports:
- "8000:8080"
networks:
- webnet
networks:
webnet:
- 初始化swarm
root@Server218 /h/d/g/d/httpserver# docker swarm init
Swarm initialized: current node (atiuy6c8k1qcig5w3br1bwf0n) is now a manager.
To add a worker to this swarm, run the following command:
docker swarm join \
--token SWMTKN-1-5e0sj0glbwl5w5rqytyqokeg91n7piwdyw9ik598x0poiz5s20-do445zhrdr5io93lplov42c2y \
172.31.237.68:2377
To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.
root@Server218 /h/d/g/d/httpserver#
- 发布服务到swarm中
root@Server218 /h/d/g/d/httpserver# docker stack deploy -c docker-compose.yml httpserver
Creating service httpserver_web
root@Server218 /h/d/g/d/httpserver#
root@Server218 /h/d/g/d/httpserver#
root@Server218 /h/d/g/d/httpserver#
root@Server218 /h/d/g/d/httpserver# docker stack ps httpserver
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
kvpvwptlj44c httpserver_web.1 httpserver:latest Server218 Running Running 14 seconds ago
6549g79dz5iz httpserver_web.2 httpserver:latest Server218 Running Running 14 seconds ago
xkz42mnetmws httpserver_web.3 httpserver:latest Server218 Running Running 14 seconds ago
rpziwzmpogn2 httpserver_web.4 httpserver:latest Server218 Running Running 14 seconds ago
y2kfe8bp09ld httpserver_web.5 httpserver:latest Server218 Running Running 6 seconds ago
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：a5419e3d3f9c"}~
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：f636819bd9c4"}~
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：97a6e3c9a064"}~
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：7f1b28e14970"}~
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：0ce251661188"}~
root@Server218 /h/d/g/d/httpserver# curl http://localhost:8000/ping
{"message":"当前为您服务的主机为：a5419e3d3f9c"}~
root@Server218 /h/d/g/d/httpserver#
- 实例弹性变化
具体的操作只需要修改docker-compose.yml中的replicas的数量即可。然后重新使用：
docker stack deploy -c docker-compose.yml httpserver
发布就可以了，可以看出swarm管理下的实例会进行自动的负载均衡。
- 关停服务
root@Server218 /h/d/g/d/httpserver# docker stack ls
NAME SERVICES
httpserver 1
root@Server218 /h/d/g/d/httpserver# docker stack rm httpserver
Removing service httpserver_web
Removing network httpserver_webnet
root@Server218 /h/d/g/d/httpserver# docker stack ls
NAME SERVICES
root@Server218 /h/d/g/d/httpserver#
- 关掉swarm
root@Server218 /h/d/g/d/httpserver# docker swarm leave --force
Node left the swarm.
root@Server218 /h/d/g/d/httpserver#
总结
最后，回头看看这个目录。
root@Server218 /h/d/g/d/httpserver# ls -al
total 16688
drwxr-xr-x 2 root root 4096 Oct 14 12:20 ./
drwxr-xr-x 5 root root 4096 Oct 14 12:00 ../
-rw-r--r-- 1 root root 121 Oct 14 12:11 Dockerfile
-rw-r--r-- 1 root root 362 Oct 14 12:01 Makefile
-rw-r--r-- 1 root root 682 Oct 14 12:01 app.go
-rw-r--r-- 1 root root 385 Oct 14 12:27 docker-compose.yml
-rwxr-xr-x 1 root root 17063672 Oct 14 12:03 httpserver*
root@Server218 /h/d/g/d/httpserver#
基本上，在docker中弹性部署自己的服务就结束了，基本上也能满足需要。