Periodically pull the package from the specific site and install it and update it.now this tool need the supervisor to manage the service。 this tools have integrated and tested with aws code pipeline and ECS.and can be used as CI/CD tools 自动安装部署工具，从指定的地址下载安装包，进行安装。并且如果有新版本发布，持续进行更新.需要使用supervisor进行服务的管理。本工具和aws code pipeline 和 ECS服务进行了集成。 可以构建持续集成和持续部署的简单环境。
Due to AWS Code Deploy can not support Docker currently, so I write this tools to demo the CI/CD process for AWS code pipeline + jenkins + manven + AWS ECS #####Integrate with AWS code pipeline, publish new built package and it's version information to your s3 bucket #####Fetch package from aws s3 and other URL #####Auto insatllation and start the application #####Keep tracting the installed version and new published version #####Get new version info from remote URL and compare with local record, if it is new one, download and update it #####Can intergrated with aws ECS, run within docker,this will short the period of CI/CD #####Multi threads mode, it can deliver more performance and stability
If you use AWS code pipeline. you should add one invoke step in you pipeline. you can use the doPreparePackage.py, this script is a lambda script. it will publish your built package and version infomation to your target s3 location. and then the codeautodeploy.py will continuously fetch the new version infomation from your s3 URL, install the package firstly and keepping to update the appliation if there is new version published.
sudo easy_install supervisor
(or) sudo pip install supervisor
echo_supervisord_conf > /etc/supervisord.conf
supervisord -c /etc/supervisord.conf
sudo vi /etc/supervisord.conf, append below content: where myapp is your appliation name
[program:myapp]
directory=/usr/local/myapp
command=nohup /usr/java/jdk1.8.0_101/bin/java -jar /usr/local/myapp/myapp.jar
user=root
autostart=true
autorestart=true
startsecs=10
startretries=3
stdout_logfile=/var/log/myapp-stdout.log
stderr_logfile=/var/log/myapp-stderr.log
[program:codeautodeploy]
directory=/usr/local/codeautodeploy
command=nohup /usr/bin/python /usr/local/codeautodeploy/codeautodeploy.py
user=root
autostart=true
autorestart=true
startsecs=10
startretries=3
stdout_logfile=/var/log/codeautodeploy-stdout.log
stderr_logfile=/var/log/codeautodeploy-stderr.log
supervisorctl reload
git clone https://github.com/wesley1975/codeautodeploy.git
sudo chmod a+x codeautodeploy/codeautodeploy.py
sudo mkdir /usr/local/codeautodeploy
sudo cp codeautodeploy/codeautodeploy.py /usr/local/codeautodeploy/
sudo cp codeautodeploy/codeautodeploy.cfg /usr/local/codeautodeploy/
sudo vi /usr/local/codeautodeploy/codeautodeploy.cfg
[codeautodeploy]
currentversion = 0
newpackmd5 = 0
localpackagename = EurekaServer-0.0.1-SNAPSHOT.jar
localinstallationpath = /usr/local/eurekaserver
remotepackagever = https://s3.amazonaws.com/wuwesley/flashsales/EurekaServer-0.0.1-SNAPSHOT.jar.info
remotepackageurl = https://s3.amazonaws.com/wuwesley/flashsales/EurekaServer-0.0.1-SNAPSHOT.jar
servicename = eurekaserver
serviceport = 8080
startservicecmd = sudo /usr/local/bin/supervisorctl start eurekaserver
stopservicecmd = sudo /usr/local/bin/supervisorctl stop eurekaserver
loglevel = INFO
sudo /usr/local/bin/supervisorctl start codeautodeploy
sudo /usr/local/bin/supervisorctl status codeautodeploy
sudo tail -n 10 -f /usr/local/codeautodeploy/codeautodeploy.log