The steps in Ubuntu includes:
$ sudo pacman -S pcmanfm mtpfs gvfs-mtp gvfs-gphoto2 libmtp
Edit the fuse configuration via:
$ sudo vim /etc/fuse.conf
user_allow_other
Mount the equipment to ~/mnt:
$ mkdir -p ~/mnt
$ mtpfs -o allow_other ~/mnt
Now open pcmanfm you will see the attached new android equipment.
Or now if you insert the android phone, the pcmanfm will automatically mount the equipment’s file system.
On ubuntu1404/1604, the template should enable HVM.
Its partition size should less than the primary storage size.
Password reset issue: Ubuntu1604, failed.
Ubuntu14.04 startup too slow, because of ntp, remove it sudo update-rc.d -f ntp remove
The reset-password script is located as
/etc/init.d/cloud-set-guest-password, then we could do following
configurations:
Remove the services via:
# update-rc.d -f cloud-set-guest-password remove
# cp /etc/init.d/cloud-setup-guest-password /usr/bin/
# chmod 777 /usr/bin/cloud-setup-guest-password
Before because you set update-rc.d cloud-set-guest-password defaults 98, so
first you should remove it.
Then Create a systemd service as /lib/systemd/system/cloudpassword.service:
[Unit]
Description=cloudpass container
[Service]
Type=idle
Restart=always
ExecStart=/usr/bin/cloud-set-guest-password
ExecStop=/usr/bin/echo hello
[Install]
WantedBy=multi-user.target
Enable the service via:
# systemctl enable cloudpassword
After the network is online:
# systemctl enable systemd-networkd-wait-online.service
The steps could refers to another article.
Switch to UTC time:
# cp /usr/share/zoneinfo/UTC /etc/localtime
Listening on other ports:
$ gitbook --port 14001 serve
Install and configuration steps are listed as following(Ubuntu16.04):
# apt-get install python-pip
# pip install --upgrade pip
# pip install shadowsocks
# touch /etc/shadowsocks.json
{
"server":"1xx.xxx.xxx.xxx",
"server_port":19175,
"local_address":"127.0.0.1",
"local_port":1080,
"password":"XXXXXXXXXXXXXXXXXXXXXXx",
"timeout":300,
"method":"aes-256-cfb",
"fast_open":false
}
# vim /etc/rc.local
ssserver -c /etc/shadowsocks.json -d start
Now you could continue to configure your shadowsocks client side.
Connect to existing docker-composed mariadb instance via following command:
sudo docker run -it -v /var/download/sbt/lili:/var/lib/mysql --link mariadb:mymariadb --net lilimarleen_default mariadb:latest /bin/bash
root@5d7eadc81b9e:/# mysql -h172.18.0.3 -P3306 -uroot -pexample
Welcome to the MariaDB monitor. Commands end with ; or \g.
Your MariaDB connection id is 35
Server version: 10.1.20-MariaDB-1~jessie mariadb.org binary distribution
Copyright (c) 2000, 2016, Oracle, MariaDB Corporation Ab and others.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
MariaDB [(none)]> \q
Bye
-v is for using local volume for restoring aims.--link is for linking to existing instance.-net could be get via docker net list.
For speed-up my blog, I changed the fonts/css from the template’s default
location to http://www.bootcdn.cn/?, this cdn’s speed is pretty good, the
speed has been greatly improved.
First use git log for getting all of the commit history, then you could get
the md5 for each commitment, git diff md51 md52 you could easily get the
differences.
When meeting guest has not initialize the display (yet), do following:
$ sudo qemu-system-x86_64 -net nic -net user,hostfwd=tcp::2222-:22 -hda ./test.qcow2 -boot d -cdrom /mnt/CentOS-7-x86_64-Everything-1611.iso -m 2048 --enable-kvm -vga virtio
-vga virtio solves this issue.
Run bashrc at ArchLinux:
$ vim ~/.bash_profile
if [ -f ~/.bashrc ]; then
. ~/.bashrc
fi
$ vim ~/.bashrc
Then all of the items in .bashrc could be executed.
Add following lines in to ~/.bashrc:
HISTSIZE=9000
HISTFILESIZE=$HISTSIZE
HISTCONTROL=ignorespace:ignoredups
history() {
_bash_history_sync
builtin history "$@"
}
_bash_history_sync() {
builtin history -a #1
HISTFILESIZE=$HISTSIZE #2
builtin history -c #3
builtin history -r #4
}
PROMPT_COMMAND=_bash_history_sync
Issue:
when I start a virtual network named 'default' (created by libvirt), it occur that:
"Error starting network 'default': internal error: Failed to initialize a valid firewall backend".
Solution:
$ sudo pacman -S ebtables dnsmasq
$ sudo systemctl restart libvirtd
then you could get your libvirtd networking work properly.
https://fedoraproject.org/wiki/Networking/Bridging
Manually enable the vnc-server service, then you could use vncviewer for connecting the vnc server in ArchLinux:
# firewall-cmd --zone=public --add-service=vnc-server
Related commands:
List services:
# firewall-cmd --zone=public --list-services
Get all of the listed services:
# firewall-cmd --get-services
Get all of the zones:
# firewall-cmd --get-active-zones
# firewall-cmd --get-zones
# firewall-cmd --get-default-zone
Execute like following:
mysql -u $user -p$passsword -Bse "command1;command2;....;commandn"
Generate md5 encrypted passwd using passlib:
$ sudo pip2 install passlib
$ python2 -c "from passlib.hash import sha512_crypt; import getpass; print sha512_crypt.encrypt(getpass.getpass())"
In archlinux, do following:
$ sudo pacman -S nginx
$ sudo vim /etc/nginx/nginx.conf
location / {
root /media/sdb;
index index.html index.htm;
autoindex on;
}
$ sudo systemctl enable nginx
$ sudo systemctl start nginx
Then you could see the nginx running on your server.
Install and running minikube via:
$ yaourt minikube
$ wget https://github.com/dhiltgen/docker-machine-kvm/releases/download/v0.7.0/docker-machine-driver-kvm
$ sudo mv docker-machine-driver-kvm /usr/local/bin
$ sudo chmod 777 /usr/local/bin/*
$ sudo minikube start --vm-driver kvm
$ curl -Lo kubectl https://storage.googleapis.com/kubernetes-release/release/v1.5.3/bin/linux/amd64/kubectl && chmod +x kubectl && sudo mv kubectl /usr/local/bin/
Then you could begin using minikube.
When using hugo for building, the travis-ci complains:
enc.SetIndent undefined
Solved via edit the .travis.yml file
go:
- - 1.6
+ - 1.7
Dependency on golang 1.7 solves the problem.
Start the default network for virsh via:
$ sudo virsh net-start default
Install fonts via:
$ sudo pacman -S adobe-source-han-sans-cn-fonts adobe-source-han-sans-tw-fonts
And mac fonts via:
$ yaourt ttf-mac-fonts
Detect Windows via:
$ sudo apt-get install os-prober
$ sudo os-prober
$ sudo update-grub
UPgrade the shadowsocks, so you could get more powerful tools..
$ pip install --upgrade pi
$ sudo pip install --upgrade shadowsocks
The configuration of vagrant ssh could be viewd via:
$ vagrant ssh-config
Add this service named iptableslast.service, enable and start the service,
then system will run iptables once for setup the rules.
[Unit]
Description=iptables last rule
[Service]
Type=idle
Restart=on-failure
ExecStart=/sbin/iptables -P FORWARD ACCEPT
ExecStop=/usr/bin/echo hello
[Install]
WantedBy=multi-user.target
Specified the version of plugins:
vagrant plugin install --plugin-version 0.0.37 vagrant-libvirt
Enable it via following commands:
[root@dlp ~]# vi /etc/modprobe.d/kvm-nested.conf
# create new
options kvm_intel nested=1
[root@dlp ~]# modprobe -r kvm_intel # unload
[root@dlp ~]# modprobe kvm_intel # reload again
[root@dlp ~]# cat /sys/module/kvm_intel/parameters/nested
Y# just enabled
The eth1 won’t be startup, so manually execute service network restart in
CentOS7.
https://gist.github.com/hydra1983/22b2bed38b4f5f56caa87c830c96378d
Commandline:
$ ./docker_images.sh save_images
$ ./docker_images.sh load_images
yum download the packages src via:
# yum install yum-utils
# yumdownloader --source mypkg
Lizcat, since the centos 6 won’t support docker well, switches to Ubuntu
# wget -O /etc/yum.repos.d/epel.repo http://mirrors.aliyun.com/repo/epel-6.repo
# yum install -y python-pip wget
via yaourt downgrade, then downgrade the package name.
clearing the cookie, then you could login, or you won’t be logined into the system.
Via following steps:
解压haroopad-v0.13.1-x64.tar.gz
cd /usr/local/src
sudo mkdir haroopad
sudo tar -zxvf haroopad-v0.13.1-x64.tar.gz -C haroopad
之后会出现如下内容
control.tar.gz
data.tar.gz
debian-binary
解压 data.tar.gz，并将其子目录下边的内容拷贝到 “/” 目录下
tar zxvf data.tar.gz
sudo cp -r ./usr /
解压control.tar.gz，并赋予755权限
tar zxf control.tar.gz
chmod 755 postinst
sudo ./postinst
Simply try hostname YourNewHostName, it could immediately change your
hostname, without a restart.
Your should use parted rather than fdisk, cause parted could support gpt mode.
# parted /dev/sdb #使用parted来对GPT磁盘操作，进入交互式模式
(parted) mklabel gpt # 将MBR磁盘格式化为GPT
(parted) mkpart primary ext4 0% 100% 将所有容量分为一个主分区
(parted) p #打印当前分区
(parted) q #退出
On pdksh missing.
Solution
Solution for 11.2.0.3 64-bit (x86-64)
If you have downloaded the 11.2.0.3 media from My Oracle Support (MOS) and extracted the software to <path>/database, do the following:
Change directory to <path>/database/stage/cvu/cv/admin
Backup cvu_config cp cvu_config backup_cvu_config
Edit cvu_config and change the following line CV_ASSUME_DISTID=OEL4 to CV_ASSUME_DISTID=OEL6
Save the updated cvu_config file
Install the 11.2.0.3 software using:
<path>/database/runInstaller
cd <path>/database
./runInstaller
OUI should now perform the OEL6 prerequisite checks (which are identical to the RHEL6 prerequisite checks) and no longer report that packages elfutils-libelf-devel-0.97 and pdksh-5.2.14 are missing
Changed from
register: check_powershell5
when: "{{ ansible_PSVersionTable.Major|int < 5 }}"
tags: win_powershell
to
register: check_powershell5
when: ansible_PSVersionTable.Major|int < 5
tags: win_powershell
could solve the jinjia problem, when statements should not include jinja2 templating delimiters such as {{ }} or {% %}.
For building larger docker image, you have to enable dm.basesize, however
the docker’s new version have to specify more parameters, the example is
listed as:
ExecStart=/usr/bin/dockerd -s devicemapper --storage-opt dm.basesize=50G --registry-mirror=http://1a653205.m.daocloud.io -H fd://
Add samba user and set its username/passwd via:
# useradd xxx
# smbpasswd -a xxx
Add following definition into the vagrantfile:
config.ssh.private_key_path = "/home/joseph/.ssh/id_rsa"
When the runner is not set, do enable this manually for project:
You could do via following ways:
cd src/
tar czf ${topdir}/SOURCES/${PKGNAME}.tar.gz $1
cd ..
Or:
tar czf ${topdir}/SOURCES/${PKGNAME}.tar.gz -C src/ .
Using following command:
# 7z a -tzip -v1024m secondnode
Recover:
# 7z e xxx.zip.001
hald introduction:
https://systembash.com/disabling-the-hald-addon-storage-service-on-centosredhat/
The hald – Hardware Access Layer Daemon – runs several processes in order to keep track of what hardware is installed on your system. This includes polling USB Drives and ‘hot-swap’ devices to check for changes along with a host of other tasks.
You might see it running on your system as follows:
2474 ? S 0:00 \_ hald-runner
2481 ? S 0:00 \_ hald-addon-acpi: listening on acpid socket /var/run/acpid.socket
2487 ? S 0:00 \_ hald-addon-keyboard: listening on /dev/input/event0
2495 ? S 41:47 \_ hald-addon-storage: polling /dev/hdc
If your system is static and the devices do not change, you can actually disable this service using a policy entry.
Create a file in your policy directory, for example /etc/hal/fdi/policy/99-custom.fdi. Add the text:
Solved(completely disable hald):
https://zhuqy.wordpress.com/2011/04/07/disable-the-hald-addon-storage-completely/
<?xml version="1.0" encoding="UTF-8"?>
<deviceinfo version="0.2">
<device>
<match key="storage.removable" bool="true">
<remove key="info.addons" type="strlist">hald-addon-storage</remove>
</match>
</device>
</deviceinfo>
After created vagrant, use ansible the following way:
$ ansible --private-key=~/.vagrant.d/insecure_private_key -u vagrant ...
for testing module ping/command, etc.
via apt-get clean, then you could get the clean directory.
Startup:
https://juejin.im/entry/588940fc2f301e0069b2397d
Add following items in your ~/.zshrc, then cowsay will always say blablabla:
if [ -x /usr/bin/cowsay -a -x /usr/bin/fortune ]; then
fortune | cowsay
fi
Good software for drawing pictures.
http://dockone.io/article/2495
http://blog.csdn.net/beckdon/article/details/36174925
use last reboot for checking the reboot reason.
use cat /var/log/messages | more for knowing the detailed reason.
Previously I found my virsh network default won’t startup automatically, so I
add one tips in ~/.config/awesome/rc.lua, but yesterday I found one tip for
adding the default network auto-start:
$ sudo virsh net-autostart default
So from now on my default networking for libvirt will start automatically every time.
I changed my mycapscr to following, so I could save 3 seconds for each
screenshot.
scrot -s '%Y_%m_%d_%H_%M_%S_$wx$h.jpg' -e 'mv $f ~/capscr/'
ls -lt ~/capscr | sed -n 2p | awk {'print $9'} | xclip -r
First you have to find the vmdk files of vagrant, then convert it from vmdk to qcow2:
$ sudo qemu-img convert -f vmdk -O qcow2 box-disk001.vmdk gitlab.qcow2
The gitlab.qcow2 is the one you want, enjoy it.
Install your own dce:
# gzip -dc dce-2.6.0.tar.gz | docker load
# bash -c "$(docker run --rm daocloud.io/daocloud/dce:2.6.0 install)"
Press ctrl+w, then J
First you should use apt-cache policy xxxx to get all of the possible
version of this software, then, you specify the version in following command
like:
# sudo apt-get install gparted=0.16.1-1
In /usr/libexec/qemu-kvm.
via nmtui in centos7 you could setup the hostname in centos7.
Via virsh net-dumpxml>abc.xml ; vim abc.xml; virsh net-define abc.xml, you
could edit your xml like following:
<ip address="192.168.122.1" netmask="255.255.255.0">
<dhcp>
<range start="192.168.122.100" end="192.168.122.254" />
<host mac="00:16:3e:77:e2:ed" name="foo.example.com" ip="192.168.122.10" />
<host mac="00:16:3e:3e:a9:1a" name="bar.example.com" ip="192.168.122.11" />
</dhcp>
</ip>
Set the parameter of docker via following command:
$ sudo ros config get rancher.docker.insecure_registry
$ sudo ros config set rancher.docker.insecure_registry ["192.168.124.211:8088"]
more detailed parameters could be found at:
http://rancher.com/docs/os/v1.0/en/configuration/docker/
In software testing, a canary is a push of programming code changes to a small group of end users who are unaware that they are receiving new code. Because the canary is only distributed to a small number of users, its impact is relatively small and changes can be reversed quickly should the new code prove to be buggy. Canary tests, which are often automated, are run after testing in a sandbox environment has been completed.
Real meaning:
煤矿工人过去带着金丝雀下井。这种鸟对危险气体的敏感度超过人。如果金丝雀死了，矿工便知道井下有危险气体，需要撤离。」
原来 Canary build 是这个意思
Directly copy from container to host via:
# docker cp 0a63783b716a:/root/d /tmp/
When some window displayed error, using 24-bit color.
For using e-sata and udiskie, you have to change the following for using properly:
➜ rules.d pwd
/etc/udev/rules.d
➜ rules.d cat 99-udisks2.rules
# UDISKS_FILESYSTEM_SHARED
# ==1: mount filesystem to a shared directory (/media/VolumeName)
# ==0: mount filesystem to a private directory (/run/media/$USER/VolumeName)
# See udisks(8)
ENV{ID_FS_USAGE}=="filesystem|other|crypto", ENV{UDISKS_FILESYSTEM_SHARED}="1"
Thus your disk will be mounted under in /media rather than in
/run/media/xxx, this means your directoy will be seemed as shared, so your
libvirt daemon will access it properly.
dashbaord v1.6.2 has a bug that you could not get heapster info.
t is solved. This is a cluster issue. In 1.6.2 dashboard was building wrong path to heapster and there was always the same error that health check to heapster was failing because dashboard could not access it.
Change new line from unix to windows:
# unix2dos xxxx
Change filename from UTF-8 to GBK:
# convmv -f zh_CN.UTF-8 -t GBK -r --notest ./
Change file content from UTF-8 to GBK:
# enca -L zh_CN -x GBK *
Refers to :
http://blog.csdn.net/a280606790/article/details/8504133
When changing ip in gitlab server, do following steps:
The correct place in an Omnibus install is:
/etc/gitlab/gitlab.rb
external_url 'http://gitlab.example.com'
Finally, you'll need to execute sudo gitlab-ctl reconfigure and sudo gitlab-ctl restart so the changes apply.
Read from the file, and generate the bat file with each person’s own ip address setting.
#!/bin/sh
while read line
do
#echo $line
NAME=`echo $line|awk {'print $1'}`
# echo $NAME
IP=`echo $line|awk {'print $4'}`
# echo $IP
ORG=`echo $line|awk {'print $2'}`
# echo $ORG
echo -n 'netsh interface ipv4 delete dnsservers "本地连接" all'>${NAME}_${ORG}.bat
echo 'netsh interface ipv4 set address name="本地连接" static 192.168.103.'${IP} '255.255.255.0 192.168.103.1'>>${NAME}_${ORG}.bat
done<inhuajing.txt
View windows uptime via:
systeminfo
Via following commands:
# timedatectl set-timezone America/Chicago
Verify new settings by typing the following two commands:
# date
# ls -l /etc/localtime
In host machine, do following, then you could forwarding the traffic from host machine 2222 to vm(192.168.122.203)‘s 22 port.
$sudo iptables -t nat -A PREROUTING -p tcp --dport 2222 -j DNAT --to-destination 192.168.122.203:22
$sudo iptables -t nat -A POSTROUTING -p tcp --dport 22 -d 192.168.122.203 -j SNAT --to 192.168.122.1
Install htpasswd in archlinux via yaourt apache-tools.
Via following command:
https://bbs.archlinux.org/viewtopic.php?id=229502
git file too large
For last 3 days I was busy designing the website for a company, I run everything in docker, and finish the design very fast. Now I want to migrate it from the old vps to a new vps, following records all of the steps.
Examine the data volume via following commands:
# docker volume ls | grep db
local makeclean_db_data
# docker volume ls | grep html
local makeclean_html
Get the volume mount point:
# docker volume inspect makeclean_html | grep -i mountpoint
"Mountpoint": "/var/lib/docker/volumes/makeclean_html/_data",
# docker volume inspect makeclean_db_data | grep -i mountpoint
"Mountpoint": "/var/lib/docker/volumes/makeclean_db_data/_data",
Now zip all of the data volume:
# tar czf db_data.tar.gz /var/lib/docker/volumes/makeclean_db_data/_data
# tar czf html.tar.gz /var/lib/docker/volumes/makeclean_html/_data
Copy the 2 gz files to your new vps.
Start the docker-compose like the steps in last article, then manually stop the docker instance, replace the data volume using the 2 gz files, start the docker instance again.
Now you will be leading the old domain name, we need to change the options of wordpress database, we create a new mysql docker instance for changing the wordpress database:
# docker run -it --link wordpress_db_1:mysql --net wordpress_default mysql:5.7 /bin/bash
root@95f633180461:/#
In this newly created docker instance you could use mysql client tools for
changing wordpress options(172.19.0.2 is your mysql server):
# mysql -h172.19.0.2 -P3306 -uroot -pwordpress
Mysql adjust commands ls listed as following:
use wordpress;
UPDATE wp_options SET option_value = replace(option_value, 'xxx.xxx.222.222:8000','localhost:8000') ;
UPDATE wp_posts SET post_content = replace(post_content, 'xxx.xxx.222.222:8000','localhost:8000') ;
UPDATE wp_comments SET comment_content = replace(comment_content, 'xxx.xxx.222.222:8000', 'localhost:8000') ;
UPDATE wp_comments SET comment_author_url = replace(comment_author_url, 'xxx.xxx.222.222:8000', 'localhost:8000') ;
\q
Now restart the docker instance, you will get the wordpress running in new vps.
Recently I want to setup a wordpress based website and quickly adjust the content, since the time is so limted, I choose docker for development and finish the task soonly, this article records all of the detailed steps.
We use docker-compose for composing the wordpress based apps, the commands listed as following:
$ mkdir -p ~/code/wordpress
$ vim docker-compose.yml
The docker-compose.yml file is listed as following:
version: '2'
services:
db:
image: mysql:5.7
volumes:
- db_data:/var/lib/mysql
restart: always
environment:
MYSQL_ROOT_PASSWORD: wordpress
MYSQL_DATABASE: wordpress
MYSQL_USER: wordpress
MYSQL_PASSWORD: wordpress
wordpress:
depends_on:
- db
image: wordpress:latest
volumes:
- html:/var/www/html
ports:
- "8000:80"
restart: always
environment:
WORDPRESS_DB_HOST: db:3306
WORDPRESS_DB_PASSWORD: wordpress
volumes:
db_data:
html:
Now use following commands for startup the wordpress or delete all of the content:
#### Startup the website
$ sudo docker-compose up -d
#### Down the wordpress
$ sudo docker-compose down
Now pick up your browser for visting http://YourIPAddress:8000 you could get
the wordpress installed and run.
We want to use the makeclean template for setting up our website(because I
am setting up a clean corporation website for my friends). Download it
directly from internet:
$ wget http://dlw.press/themes/corporate/makeclean.zip
$ mkdir -p ~/code/wordpress/makeclean
$ cp makeclean.zip ~/code/wordpress/makeclean
$ cd ~/code/wordpress/makeclean && unzip makeclean.zip
Inspect the volumes and install the makeclean template:
$ sudo docker volume ls | grep html
local makeclean_html
$ sudo docker volume inspect makeclean_html
[
{
"Name": "makeclean_html",
"Driver": "local",
"Mountpoint": "/var/lib/docker/volumes/makeclean_html/_data",
"Labels": null
}
]
Now copy the makeclean template into the docker volume:
# cp -r ~/code/wordpress/makeclean/ /var/lib/docker/volumes/makeclean_html/_data/wp-content/themes/
Now Open themes, you could see the Make Clean template has been installed:
Click Enable you could startup the template, then click Begin installing plugins for installing the plugins:
Install the following plugins:
Press One Click Demo Data for enable the demo:
Click 02 Home2 and check all of the options and begin install, you should
take for a while for waiting all of the datas being downloaded into local
directory:
Be sure to activate the Envato Toolkit plugins.
Now refresh the website’s first page you will see the good-looking website has been set.
《Docker基础与实战》第八章 (韩) 李在弘。
本机, ArchLinux.
服务器, DigitalOcean, Ubuntu14.04.
之所以选择DO的主机，是因为它位于墙外，不会有防火墙的阻挡，每次都能编译成功。
ArchLinux上安装sudo pacman -S git， 之后可以配置global:
$ git config --global user.name xxxx
$ git config --global user.email xxxx@gamil.com
用Node.js编写一个简单的Web服务器，返回Hello Docker:
var express = require('express');
var app = express();
app.get(['/', '/index.html'], function (req, res) {
res.send('Hello Docker1');
});
app.listen(80);
用于描述运行该程序的依赖关系:
{
"name": "exampleapp",
"description": "Hello Docker",
"version": "0.0.1",
"dependencies": {
"express": "4.4.x"
}
}
用于运行该APP的Docker容器定义如下:
FROM ubuntu:14.04
RUN apt-get update
RUN apt-get install -y nodejs npm
ADD app.js /var/www/app.js
ADD package.json /var/www/package.json
WORKDIR /var/www
RUN npm install
CMD nodejs app.js
需要安装Docker， Docker的安装方法不再重复，安装git, 安装方法也不再重复.
创建服务器端的git仓库:
$ cd ~
$ git init exampleapp
$ cd exampleapp
$ git config receive.denycurrentbranch ignore
配置成receive.denycurrentbranch ignore是为了便于从开发者PC接收推送的源代码.
在~/exampleapp/.git/hooks目录下，创建一个post-receive文件，内容如下:
#!/bin/bash
APP_NAME=exampleapp
APP_DIR=$HOME/$APP_NAME
REVISION=$(expr substr $(git rev-parse --verify HEAD) 1 7)
GIT_WORK_TREE=$APP_DIR git checkout -f
cd $APP_DIR
docker build --tag $APP_NAME:$REVISION .
docker stop $APP_NAME
docker rm $APP_NAME
docker run -d --name $APP_NAME -p 8068:80 $APP_NAME:$REVISION
保存并设置可执行权限:
$ chmod +x ~/exampleapp/.git/hooks/post-receive
在开发机的目录下，添加remote分支:
$ git remote add origin dash@xxx.xxx.xxx.x:exampleapp
如果你的DO上主机使用别的端口，例如8894端口，则编辑ssh的配置，为主机指定端口即可:
$ vim ~/.ssh/config
host 1xx.xx.xx.xxx
port 8894
此时每次提交的代码都会触发DO上主机的自动化构建过程，且自动侦听8068端口。
只需要更改post-receive文件即可实现自动推送流程:
#!/bin/bash
APP_NAME=exampleapp
APP_DIR=$HOME/$APP_NAME
REVISION=$(expr substr $(git rev-parse --verify HEAD) 1 7)
+++ REGISTRY=192.168.0.40:5000
+++ APP_SERVERS=(
+++ "pyrasis@192.168.0.101"
+++ "pyrasis@192.168.0.102"
+++ )
GIT_WORK_TREE=$APP_DIR git checkout -f
cd $APP_DIR
docker build --tag $APP_NAME:$REVISION .
docker tag $APP_NAME:$REVISION $REGISTRY/$APP_NAME:$REVISION
docker push $REGISTRY/$APP_NAME:$REVISION
+++ SSH="ssh -o StrictHostKeyChecking=no"
+++ for SERVER in ${APP_SERVERS[@]}
+++ do
+++ $SSH $SERVER docker pull $REGISTRY/$APP_NAME:$REVISION
+++ $SSH $SERVER docker stop $APP_NAME
+++ $SSH $SERVER docker rm $APP_NAME
+++ $SSH $SERVER docker run -d --name $APP_NAME \
+++ -p 80:80 $REGISTRY/$APP_NAME:$REVISION
+++ done
上面的代码需要事先实现中转机(即运行git的机器)与目标服务器的ssh免登录设置。