NOTES: epel has another package called docker (which is a KDE docking application and is not related to this). So, you will get an error performing these steps, if you have that package already installed.
UPDATE-1: Docker version-0.6.1 is released. Upgrade instructions are available here
UPDATE-2: Docker version-0.6.2 is released.
Docker.io is the lightweight "container engine + image repository" built on top of LXC (linux lightweight containers) to ship any application.
At the core of docker.io there are
Follow these instructions to get docker.io installed on your system.
0. Install centos 6.x
1. Disable selinux as it interferes with functionality of LXC
3. Setup hop5.in repository
4. Install docker-io package
5. See the package is installed correctly
6. Add cgroup filesystem to /etc/fstab , so that docker will work correctly
Per "Peter Zimmerman" a reboot is needed to actually mount /sys/fs/cgroup (due to the kernel doesn't have cgroup support). So, the above mount step is not really needed. As its used only upon reboot.
7. Reboot the system ; make sure to select "3.10.5-3.el6.x86_64" kernel version while booting.
8. Once system is up and running, make sure you are on right kernel
10. Print hello world from a centos 6.4 container.
11. Ping from inside the container
CREDITS: Thanks to sciurus for doing the hardwork in creating the spec file for kernel-ml-aufs.
UPDATE-1: Docker version-0.6.1 is released. Upgrade instructions are available here
UPDATE-2: Docker version-0.6.2 is released.
Docker.io is the lightweight "container engine + image repository" built on top of LXC (linux lightweight containers) to ship any application.
At the core of docker.io there are
- an application "/usr/bin/docker"
- set of pre-created operating-system/application images hosted with "docker.io"
- AUFS (Another Union filesystem) to take care of snapshots, diffs, ro/rw requirements
- LXC (Linux lightweight containers)
- cgroups (for resource management and namespacing)
Follow these instructions to get docker.io installed on your system.
0. Install centos 6.x
1. Disable selinux as it interferes with functionality of LXC
[root@localhost ~]# cat /etc/selinux/config # This file controls the state of SELinux on the system. # SELINUX= can take one of these three values: # enforcing - SELinux security policy is enforced. # permissive - SELinux prints warnings instead of enforcing. # disabled - SELinux is fully disabled. SELINUX=disabled # SELINUXTYPE= type of policy in use. Possible values are: # targeted - Only targeted network daemons are protected. # strict - Full SELinux protection. SELINUXTYPE=targeted2. Download and setup Fedora EPEL Repository
sudo yum install http://ftp.riken.jp/Linux/fedora/epel/6/i386/epel-release-6-8.noarch.rpm
3. Setup hop5.in repository
cd /etc/yum.repos.d sudo wget http://www.hop5.in/yum/el6/hop5.repo
4. Install docker-io package
[root@localhost ~]# yum install docker-io Loaded plugins: fastestmirror, security Loading mirror speeds from cached hostfile * base: centos.excellmedia.net * epel: kartolo.sby.datautama.net.id * extras: centos.excellmedia.net * updates: centos.excellmedia.net Setting up Install Process Resolving Dependencies --> Running transaction check ---> Package docker-io.x86_64 0:0.5.3-4.el6 will be installed --> Processing Dependency: lxc >= 0.8.0 for package: docker-io-0.5.3-4.el6.x86_64 --> Processing Dependency: kernel-ml-aufs >= 3.10.5 for package: docker-io-0.5.3-4.el6.x86_64 --> Running transaction check ---> Package kernel-ml-aufs.x86_64 0:3.10.5-3.el6 will be installed ---> Package lxc.x86_64 0:0.8.0-3.el6 will be installed --> Processing Dependency: liblxc.so.0()(64bit) for package: lxc-0.8.0-3.el6.x86_64 --> Running transaction check ---> Package lxc-libs.x86_64 0:0.8.0-3.el6 will be installed --> Finished Dependency Resolution Dependencies Resolved ============================================================================================ Package Arch Version Repository Size ============================================================================================ Installing: docker-io x86_64 0.5.3-4.el6 hop5 1.2 M Installing for dependencies: kernel-ml-aufs x86_64 3.10.5-3.el6 hop5 33 M lxc x86_64 0.8.0-3.el6 hop5 81 k lxc-libs x86_64 0.8.0-3.el6 hop5 75 k Transaction Summary ============================================================================================ Install 4 Package(s) Total download size: 34 M Installed size: 159 M Is this ok [y/N]: y Downloading Packages: (1/4): docker-io-0.5.3-4.el6.x86_64.rpm | 1.2 MB 00:03 (2/4): kernel-ml-aufs-3.10.5-3.el6.x86_64.rpm | 33 MB 00:33 (3/4): lxc-0.8.0-3.el6.x86_64.rpm | 81 kB 00:00 (4/4): lxc-libs-0.8.0-3.el6.x86_64.rpm | 75 kB 00:00 -------------------------------------------------------------------------------------------- Total 896 kB/s | 34 MB 00:39 Running rpm_check_debug Running Transaction Test Transaction Test Succeeded Running Transaction Installing : lxc-libs-0.8.0-3.el6.x86_64 1/4 Installing : lxc-0.8.0-3.el6.x86_64 2/4 Installing : kernel-ml-aufs-3.10.5-3.el6.x86_64 3/4 Installing : docker-io-0.5.3-4.el6.x86_64 4/4 Verifying : kernel-ml-aufs-3.10.5-3.el6.x86_64 1/4 Verifying : lxc-0.8.0-3.el6.x86_64 2/4 Verifying : lxc-libs-0.8.0-3.el6.x86_64 3/4 Verifying : docker-io-0.5.3-4.el6.x86_64 4/4 Installed: docker-io.x86_64 0:0.5.3-4.el6 Dependency Installed: kernel-ml-aufs.x86_64 0:3.10.5-3.el6 lxc.x86_64 0:0.8.0-3.el6 lxc-libs.x86_64 0:0.8.0-3.el6 Complete! [root@localhost ~]#
5. See the package is installed correctly
[root@localhost ~]# docker -h Usage of docker: -D=false: Debug mode -H=[unix:///var/run/docker.sock]: tcp://host:port to bind/connect to or unix://path/to/socket to use -api-enable-cors=false: Enable CORS requests in the remote api. -b="": Attach containers to a pre-existing network bridge. Use 'none' to disable container networking -d=false: Daemon mode -dns="": Set custom dns servers -g="/var/lib/docker": Path to graph storage base dir. -p="/var/run/docker.pid": File containing process PID -r=false: Restart previously running containers
6. Add cgroup filesystem to /etc/fstab , so that docker will work correctly
[root@localhost ~]# echo "none /sys/fs/cgroup cgroup defaults 0 0" >> /etc/fstab
[root@localhost ~]# mount /sys/fs/cgroup
Per "Peter Zimmerman" a reboot is needed to actually mount /sys/fs/cgroup (due to the kernel doesn't have cgroup support). So, the above mount step is not really needed. As its used only upon reboot.
7. Reboot the system ; make sure to select "3.10.5-3.el6.x86_64" kernel version while booting.
8. Once system is up and running, make sure you are on right kernel
[root@localhost ~]# uname -r 3.10.5-3.el6.x86_64 [root@localhost ~]# grep aufs /proc/filesystems nodev aufs9. Start the docker.io in daemon mode (in a different terminal)
[root@localhost ~]# docker -d 2013/08/21 07:47:07 WARNING: Your kernel does not support cgroup swap limit. 2013/08/21 07:47:07 Listening for HTTP on /var/run/docker.sock (unix)
10. Print hello world from a centos 6.4 container.
[root@localhost ~]# docker run centos:6.4 echo "hello world" 2013/08/21 07:48:41 POST /v1.4/containers/create 2013/08/21 07:48:41 POST /v1.4/containers/c6bc9e80097e/start 2013/08/21 07:48:41 POST /v1.4/containers/c6bc9e80097e/attach?logs=1&stderr=1&stdout=1&stream=1 hello world
11. Ping from inside the container
[root@localhost ~]# docker -dns '8.8.8.8' run centos:6.4 ping -c 3 yahoo.com 2013/08/21 08:02:15 POST /v1.4/containers/create 2013/08/21 08:02:15 POST /v1.4/containers/c40a1244f9bc/start 2013/08/21 08:02:15 POST /v1.4/containers/c40a1244f9bc/attach?logs=1&stderr=1&stdout=1&stream=1 PING yahoo.com (98.138.253.109) 56(84) bytes of data. 64 bytes from ir1.fp.vip.ne1.yahoo.com (98.138.253.109): icmp_seq=1 ttl=48 time=323 ms 64 bytes from ir1.fp.vip.ne1.yahoo.com (98.138.253.109): icmp_seq=2 ttl=48 time=329 ms 64 bytes from ir1.fp.vip.ne1.yahoo.com (98.138.253.109): icmp_seq=3 ttl=49 time=302 ms --- yahoo.com ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2304ms rtt min/avg/max/mdev = 302.032/318.318/329.656/11.807 ms
TROUBLESHOOTING
"DNS/Networking Errors inside the docker"
[root@localhost ~]# docker -dns="8.8.8.8" run centos:6.4 yum install hiphop-php 2013/08/21 07:53:05 POST /v1.4/containers/create 2013/08/21 07:53:05 POST /v1.4/containers/6d9fef14bd1a/start 2013/08/21 07:53:05 POST /v1.4/containers/6d9fef14bd1a/attach?logs=1&stderr=1&stdout=1&stream=1 Loaded plugins: fastestmirror Error: Cannot retrieve repository metadata (repomd.xml) for repository: base. Please verify its path and try again Could not retrieve mirrorlist http://mirrorlist.centos.org/?release=6&arch=x86_64&repo=os error was 14: PYCURL ERROR 6 - "Couldn't resolve host 'mirrorlist.centos.org'"
Please issue the
following commands, so that docker and its environment will be reset and above issue will be taken care of.
pkill docker
iptables -t nat -F
ifconfig docker0 down
brctl delbr docker0
docker -d
following commands:
pkill docker
iptables -t nat -F
ifconfig docker0 down
brctl delbr docker0
docker -d
It will force docker to recreate the bridge and reinit all the network rules
初始化脚本一个:
!/bin/bash
### BEGIN INIT INFO
# INIT INFO
# Provides: docker
# Required-Start: networking
# Required-Stop: networking
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: docker
# Description: docker
### END INIT INFO
. /etc/init.d/functions
NAME=docker
PID_FILE="/var/run/${NAME}.pid"
DELAY=0
status_q() {
status -p $PID_FILE $NAME > /dev/null
return $?
}
start_app() {
nohup /usr/bin/docker -d >/var/log/docker.log 2>&1 </dev/null &
return $?
}
start() {
status_q; res=$?
if [ $res -eq 3 ]; then
echo -n $"$NAME starting..."
start_app
echo $! > $PID_FILE
[ $DELAY -gt 0 ] && sleep $DELAY
status_q; res=$?
[ $res -eq 0 ] && success || failure
echo
elif [ $res -eq 0 ]; then
success; echo "$NAME is already running"
else
failure; echo "$NAME is not running, res: $res"
fi
return $res
}
stop() {
status_q; res=$?
if [ $res -eq 3 ]; then
success && echo "$NAME is not running"
else
killproc -p $PID_FILE; res=$?
if [ $res -eq 0 ]; then
rm -f PID_FILE
success && echo "$NAME has been stopped"
else
failure && echo "$NAME cannot be stopped ($res)"
fi
fi
return $res
}
case $1 in
start)
start
exit $?
;;
stop)
stop
exit $?
;;
status)
echo -n $"$NAME running..."
status_q; res=$?
[ $res -eq 0 ] && success || failure
echo
exit $res
;;
restart)
stop
start
exit $?
;;
*)
echo "Usage: $0 {start|stop|restart|status}"
exit 2
;;
esac
其他链接:
http://qing.blog.sina.com.cn/2294942122/88ca09aa33003ydp.html